User talk:LilaTretikov (WMF)

edit
Archives
1 2 3 4 5

WMF superblocks its community

Hi,

since Erik doesn't answer, I'm now sending this remark to some other WMF officers and board members. I apologize for using your time.

I'm a crat in german wp. The so-called super-protections that Erik Möller/User:Eloquence and User:JEissfeldt (WMF) have put on our common.js on sunday, acting officially on behalf of WMF, have left some blood on the carpet. Many fellow wikipedians are upset, even those who accept the media viewer (which had been the conflict's origin). Several long-time contributors have left or stopped editing due to this. Journalists picked up the case.

Personally, I strongly protest against the WMF's action, and it's failure to communicate afterwards. Our communities are capable, and willing, to handle problems like this without office-actions.

There have been no official or private comments from WMF in the last days, so I'd like to suggest you have a look and give some response to the criticism.

(apologize again, for my translation errors)

Rfc: https://meta.wikimedia.org/w/index.php?title=Requests_for_comment/Superprotect_rights

Links to ongoing discussions in german language: [1], [2], [3]

Greetings, -MBq (talk) 20:11, 12 August 2014 (UTC)

Hi, i second MBq request and especially this post by Rich. This issue is not taken lightly especially among german wikipedians. Regards, Ca$e (talk) 20:52, 12 August 2014 (UTC)

This is a temporary measure to prevent churn on the file for the lack of a better process currently in place (more on this here). LilaTretikov (talk) 21:19, 12 August 2014 (UTC)

Dear Lila, please stay that sober and unintimidated. I wholeheartedly agree with such a minimal invasion. Instead of blocking or deadminining an entire camp of notorious know-betters, just one template was commensurately frozen from those administrators' access for them to cool down and stop them from disrupting the public user experience with their edit wars. Commentor 188.61.148.188 summarized it well on their central protest page:

Funny: Someone at the Foundation is passing the same medicine to the German admins that they have been passing their authors for years. And now the German admins react with insult since they realized how it feels to be given such medicine. The San Franciscan physician has thus contributed much to the climate and future quality of the German Wikipedia, either by self-aware admins diminishing their kindergarden behaviour in the future or by at least giving their mere mortal victims another parody on the wikihierarchy to laugh about. Also very amusing how the German admins voluntarily put up squaekingly green bars on every page linking to their own idiocy.

Wishing you good luck and the proper amount of patience: IM Serious (talk) 09:27, 17 August 2014 (UTC)

Somehow liberally translated. I wrote "humiliation" ("Schmach"), not "idiocy". --188.61.148.188 14:12, 17 August 2014 (UTC)

Hello, for two years I was member of german arb-com (only just for your information).

Your post above is no answer. It maybe denote as a declaration of war. Is it that what you want? I hope not.

Please remember, that your money is earned by voluntary!

If we go away, you will earn no money! Okey, no problem for you, because you will get a new job after this...

Nearly 3 hours ago I postet that at de:wp: https://de.wikipedia.org/w/index.php?title=Wikipedia_Diskussion%3AKurier&diff=133016777&oldid=133016774 perhaps this would be a possibility to... I don`t know the right word for: "Das Gesicht wahren" --Hosse (talk) 22:28, 12 August 2014 (UTC)

Dear Hosse, in response to your suggestion, while we will not remove the software feature, we would be happy to immediately remove the protection of common.js on de.wp if there's agreement by admins that we will continue the conversation on the basis of the current state and improve it together, rather than disabling the feature. What do you think would be a reasonable way to establish that agreement? And yes, we're absolutely happy to continue the conversation on a page dedicated to this purpose. Thanks for the constructive suggestions! --LilaTretikov (talk) 23:01, 12 August 2014 (UTC)

The reasonable way is to comply to the bugreport regarding the community consensus of German Wikipedia (then also no hack is needed). You work now with and for volunteers. Noone of them has bad intentions, but the WMF lost much trust in the last two, three years for its actions. Why don't you work to convince us instead of forcing? With force you don't get anywhere in the end, you just show how weak your positions and the results of the work of the Foundation are. --Julius1990 (talk) 23:05, 12 August 2014 (UTC)

@LilaTretikov: Thank you, but no, thank you. The volunteer community have never agreed to the superprotection feature in the first place. The only real way forward for you is to unprotect MediaWiki:Common.js, take the superprotect user right from the global staff user group, and then disable this feature altogether. To have this threat of your unilateral and unlimited use of this feature hanging constantly over our heads cannot be accepted at any point in the future. odder (talk) 23:09, 12 August 2014 (UTC)

To prevent "churn on the file", direct WMF staff to quit editing it to overrule consensus, don't "superprotect" it. Seraphimblade (talk) 22:36, 12 August 2014 (UTC)

Dear Lila, I created a page at de:wp. Thank you for your nice answer. I hope you (really you) and the staff will take the chance to come into conversation with the German wikipedia. --Hosse (talk) 23:37, 12 August 2014 (UTC) PS: Sorry for my bad English

Hosse -- I will keep an eye on progress there. Looking forward to it! Thank you! LilaTretikov (talk) 00:29, 13 August 2014 (UTC)

┌─────────────────────────────────┘
Lila, "preventing churn" is a task that is dealt with daily on many Wikimedia projects, and I think it's fair to say that Wikimedians have developed sophisticated methods of doing so. Before we talk about super-protection, we should talk about normal protection for a moment, and the practices that have evolved around that. Protecting an article on Wikipedia always involves a judgment about which version is correct, and which is incorrect; therefore, there are some best practices administrators are expected to adhere to when protecting a page. I will speak only in broad strokes here, but I would encourage you to talk to some experienced administrators about this, and explore their wisdom drawn from dealing with this kind of conflict many times, in many kinds of circumstances.

There are some important principles, though, that you will surely hear about if you talk to some administrators:

1. The person protecting a page should not be involved in the dispute, and ideally should not have a very strong opinion on it at all; if the topic is Israel vs. Palestine, for instance, the ideal admin to protect a page might be a career mathematician from Kansas who has never bothered to think much about religion or the Middle East;
2. The protection is considered a temporary condition, not a decision; it is intended to encourage discussion, which is where the actual decision gets made;
3. In discussion, it is generally advised to get more uninvolved editors to give it some thought and weigh in.

These sorts of disputes are brought successfully to resolution on a daily basis throughout our projects. If they weren't, we wouldn't be a successful web site, we would be Encyclopedia Drammatica. The experience and practices that apply to resolving editorial disagreements can certainly be applied here, but they cannot be applied if the organization taking the more radical approach ("We must enable this software because we said we must enable this software") is the one applying the kind of fix that is meant to be temporary ((super)protection), and pointing to no realistic longer-term dispute resolution process on an even playing field.

If you must stick to your staff's decision, so be it; but the credibility of your organization, among a stakeholder group whose paticipation is necessary to the site's survival, is at stake. So please choose wisely. -Pete F (talk) 01:48, 13 August 2014 (UTC)

Dear Lila, please remove the superprotect feature and refrain from using it. In my humble opinion, it is not acceptable to overrule a RfC in a community-based project, especially for these reasons. Best regards, --Ghilt (talk) 07:46, 13 August 2014 (UTC)

Thanks for the details, Pete F. I am not sure the analogy is exactly applicable for software, but it is helpful nevertheless. I think the conversation we actually want to spur is specifics on what we need to have changed in the feature in question. I am asking the team to engage everyone here on user tests so we can do just that. -Lila Tretikov 13:39, 13 August 2014 (UTC)

I assume this comment above is from Lila -- I have changed the IP address to your name, please revert me if I am incorrect.

My comment was not meant to precisely correlate these admittedly different situations, but to highlight one specific practice that has been found to work very effectively in resolving disputes -- and maybe I did not make this clear enough:

Reverting something to the state it was in prior to the dispute is a tremendously powerful technique for setting the stage for dispute resolution.

With the Visual Editor, the WMF did in fact revert the feature, but as far as I know it did not take steps to continue the discussion in a more generative way. This may have been a missed opportunity. (I am not intimately familiar with how things went with VE, and I know it was before your time.)

With the Media Viewer, however, the WMF is continuing to refuse to take the one step that is guaranteed to interrupt the drama and discord: simply revert the default enabling of the software. The software doesn't have to be removed, none of the 3 projects that have had RfCs have called for its complete removal. But reversing the ill-considered decision to enable it by default on, at least, these 3 sites is a very clear precondition for more sober reflection and deliberation. And let me be clear: I am not stating a personal boundary or condition here, and I am not in a position to negotiate. I am simply stating what the clearly expressed expectation of a very large group of users is. I have no ability to change those expectations, any more than you do.

But as I said initially -- please find some Wikipedia or Commons administrators who have actively worked on dispute resolution, and ask them about these things. You needn't take my word for it -- we have a great many people in our community with deep experience in these matters, much deeper than mine. And a great many of them have not bothered to comment on the Media Viewer situation, and so they might be in a position to give you a more dispassionate and less biased opinion than I am. -Pete F (talk) 19:07, 13 August 2014 (UTC)

You and Erik claim superblocking is a temporary measure but you also name a precondition (if there's agreement by admins…). So if there is no agreement then superblocking won't be temporary, am I right? Somehow this sounds like teaching Wikipedia's community for educating the world. I cannot believe that this is of interest to the Foundation. Please remove the superprotect feature. Maybe this feature is needed but first of all we need a discussion when this feature shall be used and by whom. NNW (talk) 08:18, 13 August 2014 (UTC)

It is temporary for many reasons, including that we are planning to make process changes. But you are right, the timeline is an issue right now. We need to resolve the MV issues so we don't have this corse-grained hammer that only can do on or off switch. We need to collaborate to improve, rather then flipping switches. -- 2601:6:2080:187:A54B:B04B:FB9:1FDE 13:39, 13 August 2014 (UTC)

Dear Lila, your accountability as the Executive Director: „Your are the person ultimately responsible for the direction and actions of the WMF.“ Could you please explain what position WMF takes on the Universal Declaration on Cultural Diversity (UNESCO) – refering to WMF superblocks its community. --Edward Steintain (talk) 15:08, 13 August 2014 (UTC)

On a Scale of Billions

All,

Our site, technology and content have far reaching impact and scale. Today, we do not fully appreciate ourselves as a top-5 website from an operational standpoint. This was surprising to me and I would like to bring clarity to what this means in practice for the benefit of the staff as well as the community. I would like to outline a number of operational principles to help us navigate towards this goal. These are not something we are inventing: these are widely used high-scale product practices; we will need to modify and adopt for our use. I look forward to discussing them with you all here.

These core operational principles are the following:

• Responsibility: As a global top-5 website relied on by millions around the world, we have responsibilities and obligations to the world at large with respect to basic stewardship of the technology and brands under our care. This means we are responsible to ensure access to knowledge for every person in the world who can access the internet, any time.

• Input: The needs of billions of users are rapidly evolving. Understanding these needs requires methodology different from what we know and use today. Our current methods are no longer sufficient. We have outgrown that scale, and we require broader, ongoing testing and validation as well as deeper understanding of our users: from viewers to expert editors. We plan to develop and use improved methods over time consistent with best practices in product development. Community discussions as well as staff expertise will always be important, but not sole, inputs in determining the definition of a product and its effectiveness.

• Scope: The WMF is ultimately responsible for the security, stability, consistency, and evolution of the Wikimedia sites and brands; fundraising to achieve global reach; and specific programs to accelerate the goals of our mission.

• Operational Controls and Process: Best practices require that software and operational changes receive appropriate validation and review prior to deployment. Separation of duties, code reviews, training and validation for technical permissions as well as an unambiguous process are required for that. As the organization responsible for the site and the servers, WMF must clarify all relevant practices.

• Accountability: As the Executive Director I am the person ultimately responsible for the direction and actions of the WMF. WMF makes all decisions as a team with my oversight, and we stand by these decisions together.

• Debate, disagreement, and civility: I encourage you to discuss, and debate, any decisions as you see appropriate, and you should feel free to bring issues to my attention. However, personal attacks are never OK on our projects, email lists, discussion pages, or other digital channels. Commentary on individuals is wrong and uncivil: we should demand better of ourselves. I would like to see all comments focus on constructive ideas and suggestions, policies, responsibilities, and, most importantly, desired outcomes.

We owe the world a clear strategy, policy, process, and prioritization of the work completed by the WMF. We are committed to establishing those in the open with clear expectations consistent with our mission so we can hold ourselves responsible for our failures and successes. This process must have multiple opportunities for community feedback. We realize that has not always been the case in the past, but this will be one of my top priorities as Executive Director. You should be seeing more on those in the upcoming weeks.

LilaTretikov (talk) 21:00, 12 August 2014 (UTC)

Comments

Splitting this off for readability:

First of all, you owe everything to the editors that created the content people can now access freely. From the first dollar donation to the great amount of money the Foundation recieves now every year, it all comes in the end as the result of years of work by volunteers. You and the Foundation should finally show a little humbleness regarding this fact. That would already take most of the sharpness out of any conflict. Totalitarian actions like Erik Möller took in the media viewer case can ultimately lead to really destructive incidents. You may not fear forks. You maybe should see what a fork did to the development of the Spanish Wikipedia. You don't want to try out such effects just for going with the Foundation'S head through the wall. You lost the trust of many editors in many projects. You won't get it back fast or easily. And this statement with no sorry for the opressive actions doesn't make me trust you more. --Julius1990 (talk) 22:34, 12 August 2014 (UTC)

I agree, dear Julius1990: your labor of editing is amazing, invaluable and is the reason we are all here today. Thank you. LilaTretikov (talk) 23:05, 12 August 2014 (UTC)

Lila, me and i think most of the other editors everywhere in the projects don't want personal thanks or honor. We want to be respected by the Foundation. We should work together, but that means that you can't use force on the projects. You have no moral right to do so. Noone will oppose good features, but you should change your processes. Requests for comment/Superprotect rights lists some ideas that can work. Maybe it doesn't go then as fast as you are used to regarding the implimenattion of developed software features. But Wikipedias are no normal companies and you should respect that. You should take care of that. In the last three years all i hear is "we loose editors", "we need editors" ... and so on. Engaging new editors is fine thing, but trampling on those who already donated so much work seems to foil that. --Julius1990 (talk) 23:10, 12 August 2014 (UTC)

+1. I should add that Wikipedia is no company at all. Wikipedia is not a commodity, it is a movement led by its editors.--Aschmidt (talk) 23:19, 12 August 2014 (UTC)

I will read through that, Thank you. Our challenge is two fold: 1. we need to support many different types of users: editors that are very experienced, newbies that are currently leaving at a rate of 99+:1 because the experience is terrible, readers, searchers, embedded content, new forms of contributions... just to name a few, 2. we are already starting to loose global relevance, and that is a serious issue -- so speed is of the essence. How do we solve for those while the world is changing way faster than we are... ? LilaTretikov (talk) 23:33, 12 August 2014 (UTC)

Lila, on point #1: It is of critical importance that the WMF begin to recognize that it does not have a monopoly on this perspective. In 2009-10 we had a Strategic Planning process that involved 1000 stakeholders, who all bought into that concept; in fact, I think of the many, many the Wikimedians I have encountered, only the tiniest handful are resistent to that notion, and think only of their own convenience. This is a point repeated over and over by WMF staff, as though it were not grasped by volunteers and other stakeholders; please ponder how strange that must seem to those of us who have been volunteering for the projects specifically with the purpose of sharing knowledge with the rest of humanity for 5, 10, or more years. Please ponder how strange the idea that the WMF has greater clarity on this point than those of us who work with new contributors every day might seem to some outside the walls of WMF. -Pete F (talk) 23:47, 12 August 2014 (UTC)

And on point #2, please consider that speed is only beneficial if you are running in the right direction. If you are running backwards, speed makes the problem worse. -Pete F (talk) 23:52, 12 August 2014 (UTC)

Thanks, Pete F. On #2 -- It is true, we have to work towards the right things. Doing nothing, on the other hand, is our biggest danger. Because any progress takes trials and errors and rapid course corrections. On #1 -- we are actually formulating an ongoing community engagement around strategic and operational goals. You should see some movement on that soon. In my personal experience the best innovations come from externally generated ideas -- we just need a better listening ear and a process for prioritization. Thanks for brining it up. LilaTretikov (talk) 00:44, 13 August 2014 (UTC)

"A better listening ear and a process for prioritization" -- you and I agree 100% in principle. On #2, though -- no -- making very rapid changes to the platform, creating chaos and a sense of disrespect, is far worse than merely doing nothing. If the WMF simply keeps the servers running, and permits the volunteer community to do its thing without introducing unnecessary drama, I believe things will go fairly well. It is possible for the WMF to have a far greater impact than that, of course -- but the recent impact has generally been a negative one, not a positive one.

At any rate, thank you for the quick reply! I hope it is possible for you to put some of these principles into action, specifically on the unprecedented approach taken to overruling community consensus in favor of a very imnmature software product. -Pete F (talk) 01:03, 13 August 2014 (UTC)

@LilaTretikov: What a disappointing load of waffles. How does this relate to the superprotection feature? Are you planning to answer any of the questions that are being asked by the community, or do you think that the best way forward is for the Foundation to keep ignoring our questions and treat us with this sort of rubbishy speeches? odder (talk) 22:35, 12 August 2014 (UTC)

Dear odder -- it was not meant to. Please see the thread above for responses on that issue. Thank you! LilaTretikov (talk) 23:05, 12 August 2014 (UTC)

Your 'Accountability' paragraph: It would be much better if the foundation saw volunteers as part of the team. See my MV Arbcom case input for background. Regards, Ariconte (talk) 23:09, 12 August 2014 (UTC)

Ariconte, agreed. Could you please link to the case? Thanks! LilaTretikov (talk) 23:24, 12 August 2014 (UTC

https://en.wikipedia.org/wiki/Wikipedia:Arbitration/Requests/Case/Media_Viewer_RfC/Evidence#Evidence_presented_by_Ariconte and https://en.wikipedia.org/wiki/Wikipedia:Arbitration/Requests/Case/Media_Viewer_RfC/Workshop#Proposals_by_User:Ariconte Regards, Ariconte (talk) 00:56, 13 August 2014 (UTC)

Lila, your statement is simply embarassing. It is splitting the WMF from its editors, while it were the latter that have made Wikipedia what it is today. This has led to a great number of regular editors on German Wikipedia turning away from the project and going inactive. May I also remind you that you and the rest of the Wikimedia staff are in fact our employees? You live on the donor's money that the editors have made the donors to give to us as an entity. This is our project, and we do not intend to let it be taken away from us by any action. You are in fact obliged to look for the editors in the first place. If you don't understand this really simple point you will lose most of the core of us rather soon. I have a hunch that this is exactly what you are aiming at, for whatever reason. And this is a sad situation for most of us because it puts an end to Wikipedia as we know it. If this is your mission, please say it loud and clear, so that everyone can hear it, or correct your course and remove the superprotect privilege. I for one will not contribute any more unless this new right has been removed.--Aschmidt (talk) 23:10, 12 August 2014 (UTC)

Dear Aschmidt -- our common mission is to educate the world -- that is... everyone. We all have a part in this, no? Editors like yourself are critical, but we are all here for a reason: helping others. LilaTretikov (talk) 23:24, 12 August 2014 (UTC)

Lila, we discuss the superprotect right here which has not been consented by the community. We discuss the demobilisation of regular editors your recent decisions have led to. Real leadership looks different. I am sorry to say that you are missing the point completely. You are not my ED.--Aschmidt (talk) 23:28, 12 August 2014 (UTC)

Aschmidt -- sincerely sorry to hear that. We are not discussing superprotect in this thread however, it is the thread above. We are talking about our responsibility for the generations to come in this world and how we can ensure we do the right thing for them. LilaTretikov (talk) 00:32, 13 August 2014 (UTC)

Hm, so you assign the proper place for me where to discuss which points with you? Well, that's one way to do it. I prefer another one, as do most others on this thread, as you may have noticed. We discuss how to abolish the superprotect right where we prefer. Be flexible, please, and approach the editors where they are. They are in this thread now, not the one above. I would like to give you one more example: Just some two months ago I began a project to bring the Wikipedia Library to German Wikipedia. I have already put a lot of time and effort into this. I've talked to publishers, I've talked to WMDE, and to a lot of other people to make this happen. I've been successful in this, we only talk about how to do it, not whether to do it. We wanted to start at the end of this month. But if you proceed like this there simply will be no more long-time editors on German Wikipedia left I will be able to distribute any accounts to. So I've put the project on halt. Most German editors who belong to the core of our community are upset and indignated at what the Foundation does since Saturday/Sunday. Perhaps you did not notice at all what you were doing because this is the way you've been treating English Wikipedia for quite a while already. But this won't do with us, either. Traffic on the Wikipedia Kurier talk page has risen from some 500 to 2500 visits per day. And the news has already spread to the big German IT news outlets Heise and Golem, so it has reached its an important group of donors already. You are in charge. What would you do now? Remember who is your boss? It's the editors.--Aschmidt (talk) 08:42, 13 August 2014 (UTC)

Fair enough -- feel free to discuss here -- I just could not find my way and started loosing threads. I think the spike of interest is not a bad thing -- what we need is to turn to figuring out a better process, one that is representative of all users and definitely the editors. I will give you an example: MV has been in Beta since November 2013. On DE the Beta link was removed, seemingly without code review. This was one of the reasons people in Germany may not have gotten as much exposure to MV as they otherwise would. This is a small example, but we should have caught this problem and fixed it, engaged better, possibly posted on everyone's talk pages, etc...The process needs improvement. LilaTretikov (talk) 15:51, 13 August 2014 (UTC)

Hi Lila, just to protect you from mistakes: This is not about "people in Germany" because the de:WP is a germanspeaking community. It comprises users from Austria, Swizzerland, Liechtenstein, germanspeaking minorities in Belgium, the Netherlands etc. and at last also Germany. Don´t mix that up, dangerous bug :-) --Gleiberg (talk) 16:29, 13 August 2014 (UTC)

That's right. My bad. Thanks, LilaTretikov (talk) 16:42, 13 August 2014 (UTC)

Aschmidt's words don't speak for me, at all. Nathan ^T 23:35, 12 August 2014 (UTC)

• @LilaTreitikov. : Per your statement: "our common mission is to educate the world." — That's an empty slogan. WMF's duty is to raise funds, operate the platform, efficiently manage programs, and to create the software tools that the editing community needs. The 10,000 volunteers' duty is to create content and to manage and improve the casually-contributed content of ephemeral drop-by editors, handling quality control and manipulation of content, using the best tools available to it. The 500,000,000 readers just want accurate, honest, reliable, verifiable information at their ready disposal. We're not "educating the world," we're building an encyclopedia. You create the tools; We use the tools. Carrite (talk) 01:26, 13 August 2014 (UTC)

Carrite -- this is a great note. Please see my point under Accountability above. I believe we are saying the same thing. What I was talking vis-a-vis "educating the world" is the mission of the combined effort, that spans more than encyclopedia, and includes other projects. LilaTretikov (talk)

• You seem to be suggesting that we are no longer a volunteer project. If that is so, do you realize that you potentially jeopardize the legal protection the WMF has? Either the community does stuff to the individual projects without being forced around by Moller or the Foundation can be held responsible. There is no middle ground here, and if the legal counsel at Wikimedia claims otherwise then you better find new counsel. Already, there are some cases such as the monkey case that could have some major financial ramifications, and there is a lot of money sitting around at the Foundation that a UK court, for instance, would be happy to divy up. Ottava Rima (talk) 03:43, 13 August 2014 (UTC)

No, that is not what I am suggesting. LilaTretikov (talk) 15:51, 13 August 2014 (UTC)

I do not care about the mediaviewer itself. I just wonder; how could wikipedia survive so many years and get so large and well known without the WMV superpower? How were conflicts like this solved in the past? ...Sicherlich ^Post 05:26, 13 August 2014 (UTC) but at least finally someone of the WMV is talking. Not on de with the parties acutally involved but on a private talk page but hey I'm so greatfull

I agree with Sicherlich. @LilaTretikov: a statement here (minus the PR talk) would be much appreciated -- especially since Eloquence/Möller can’t comment at the German Wikipedia at the moment. Rgds _{• • hugarheimur} 08:12, 13 August 2014 (UTC)

+1 I agree with Sicherlich. --Steinsplitter (talk) 09:10, 13 August 2014 (UTC)

@Lila I encourage you to discuss, and debate, any decisions as you see appropriate, and you should feel free to bring issues to my attention. However, personal attacks are never OK on our projects, email lists, discussion pages, or other digital channels. Commentary on individuals is wrong and uncivil: we should demand better of ourselves. I would like to see all comments focus on constructive ideas and suggestions, policies, responsibilities, and, most importantly, desired outcomes. -> Did you tell this to Erik as well? When you have someone who calls himself "Deputy Director of the Wikimedia Foundation" making threaths to volunteers, well, that is the kind of thing you need to escalate a discussion. I think that it is brave to say that you have the ultimately responsible on they other hand. But what are you going to do to prevent this kind of drama in the future? This could and should not have happened. I don't have a strong opinion about the MediaViewer for the record but things where rushed ones again. This happened with the Visual Editor, this happened now. Are you going to change the develop process? Less rushy, primary needs first (SUL-completion for example, maybe some better protection against spam bots) and than maybe a tool to make things look nicer. Another improvement would be to hire somone to support the developers in their communication since they can clearly need someone to assist them. Such a person could also help preventing Erik from doing something stupid like escalating everything when he should have deescalated stuff. Just some thoughts and hopefully you will learn from this fiasco. Natuur12 (talk) 09:50, 13 August 2014 (UTC)

Yes, yes and yes -- on the specific points, Natuur12: we need to change the development process (just starting on that, our current process is terribly immature for our scale), we need a communications person (this is in-flight), we need to clarify policies/roles (planned). Clarity around all of the above would prevent this type of things from happening in the future. LilaTretikov (talk) 15:21, 13 August 2014 (UTC)

Dear Lila, on your above post, as well as in your post to the mailing list, you are using highly irritating wordings. Here, you say "we need to clarify policies/roles" and there you speak of "clarifying" "development and deployment process including rights and responsibilities". Please be advised that, assuming by "we" you mean WMF staff/board, this is not at all to your availability. There already are conventions in place and active, governing said "rights and responsibilities", which, according to an overwhelming perception, directly exclude measures (introducting "superprotectors" without and even against community consensus and (mis)using them accordingly) such as the ones we are talking here. Several of them have already been named especially on this page. Recent WMF actions are already seen as a breach of contract towards the active contributors of content and also of the technical background. In this light, you (i.e. WMF staff) has stepped out of line and has to reconsider its very rights and responsibilities. In this regard, you should follow the steps outlined above, e.g. by Rich and odder. Afterwards, we can talk about further considerations. But not before that. Ca$e (talk) 18:05, 13 August 2014 (UTC)

I am one of those rare creatures ... a new female editor (I joined deWP in Jan 2014). Up to now I did enjoy editing and getting to know the community of Wikipedia. I met Jan Eissfeldt at a "Stammtisch" (informal Meeting) in Munich and got an idea, how complex the process of communication between communities and staff/software developers can be. Being an Idealist myself and seeing the idealistic background of WP I was confident, that all those matters could be resolved in a (perhaps sometimes chaotic) peaceful manner. At the moment I am very confused. The Mediaviewer is no issue that moves me. For my own purposes as an editor i disactivated it, but I can see, how it could be useful for other users. However it is disconcerting to me, to see WP-Colleagues deeply offended and hurt, Colleagues that I have come to trust and respect and that obviously have done a lot to build the deWP. Please come to the German Discussion Page linked above and ask other Staff-Members to contribute there, too, so that a constuctive communication can develop an the feeling of being misused by the WMF does not spread further in the deWP. Thank you very much --Kritzolina (talk) 09:48, 13 August 2014 (UTC)

Hi Kritzolina -- great to meet you. I think all these matters can and will be resolved. I will do my best to read thought the German discussion page although, to the best of mine (and Google Translate's) ability. -- LilaTretikov (talk) 15:58, 13 August 2014 (UTC)

Thanks a lot, LilaTretikov - for your answer here and for entering the active disussion on deWP. In my line of word I have seen a lot of cooperations of professionals and volunteers (on a much smaller scale of course). There is and always will be friction. Communication ... and more communication ... an more communication is the only way I know to deal with it. I am glad, you have entered into it in the name of WMF. It would be great to read other voices from the WMF as well. --Kritzolina (talk) 19:55, 13 August 2014 (UTC)

Comments (cont'd)

Dear Lila,

seconding the appreciation by Theo1011 of your direct engagement in the discussion, i nevertheless find your answers not quite satisfactory.

This starts with your remark concerning a) the use (by many seen as misuse) of developer resources for the unprecedented modification of the background-software (new „superprotect“ right) creating a new hierarchy and separation between WMF staff and directly elected maintainers of local wikipedias, b) the application of this modification to superimpose a strategic position towards the implication of a software which a majority of active contributors does not want (in this form). To this you replied:

„This is a temporary measure to prevent churn on the file for the lack of a better process currently in place (more on this here)“

I take this reply as implying your consent in using such an unprecedented measure, which many contributors regard as directly in contradiction with conventions regarding the relationship of WMF and local communities, as explicitly stated and as implicitly in action for many years. Is this understanding of your reply correct? If not, i expect you (= you in person and every other WMF staff/board member) to intervene in following the steps outlined above. Allow me to quote (from Rich):

1 Un-super-protect the page.

2 Apologise to the community

3 Give an undertaking not to use superprotect except in a clear emergency (though I can't think of one that would justify its use) - or disable the feature

4 Engage with the community to discuss superprotect.

Once this discussion is concluded, then and only then should a discussion about media-viewer be entered into, meanwhile the community consensus should be respected.

Regarding your follow-up-reply:

„we would be happy to immediately remove the protection of common.js on de.wp if there's agreement by admins that we will continue the conversation on the basis of the current state and improve it together“

You again speak here as „we“, so i take it as implied that you knew and approved of the above-mentioned measures? Your condition „if there's agreement by admins“ is, however, besides the issue. First and foremost, you have to respect local consensus. Admins will and have to follow this consensus - a so-called „Meinungsbild“ in german wikipedia has a greater binding impact than an „RFC“ here; any admin in the german wikipedia who acts against a binding „Meinungsbild“ will presumably have his admin rights revoked by the german community in short time. Again, the steps you should take have been named above and also by odder.

Furthermore, as Julius1990 already pointed out, „The reasonable way is to comply to the bugreport“. Not only me, but also several active developers and experts involved in the technical background work find the closing of said bugreport highly against all governing conventions in this regard. One could indeed see this closing as the decivise provocative step having lead to the current, unacceptable situation.

As highlighted already by Rich, now is not the time to talk about the MediaViewer in itself but for you (i.e. all WMF staff) to try and repair the damage you caused. However, please keep in mind what e.g. Pete pointed out ([4]):

„many of us who have opposed the default enabling of the Media Viewer have done so *not* on the basis that we personally dislike it, but on the basis that we believe it causes problems for the process of helping readers become effective editors [...]“.

You tried to separate the above measures and their legitimation from your overall perspective on WP/WMF in general. However, as many here and elsewhere have pointed out, this idea of seperation is besides the issue, as clearly your overall perspective governed your appreciation (if i take you correctly on both your replies mentioned above) of said measures. Exactly this interconnection – and not at all the rather minor technical details – is what has produced the largely felt grave affront against local communities.

Here are a few voices from the discussion (all of them by highly active, well-respected and well-known contributors in their respective areas, among them several former members of the german arbcom, e.g.). I took the liberty to edit and partly translate them in order to keep this already large post a bit shorter, please excuse any misrepresentations thereby caused.

Pete ([5]):

„The WMF chose to "Narrow Focus" a couple years ago. I believe that what got "narrowed out" was, by and large, processes that serve the secondary purpose of helping the WMF educate itself, in an ongoing way, about how its projects and communities operate. I believe we are seeing the effects of that decision now.“

Magiers ([6], [7]):

"They are not worried about chasing the active users away. They don't want us. They want new users, different users, facebook users, hyperactive people, that make pictures anywhere and edit with their mobile, that add to the article what goes through their mind at the moment, not what is written in reputable books and found out after a long and tiring reasearch. They want masses, a movement, millions of bot-articles. And they hope that the new masses will finally chase the old users away, those who are questioning their godlike decisions and their godlike self-declared superpowers. You see this in every answer: "You, who have argued and voted, are only the old editors, but we care about the silent readers and the fictive new editors. So of course we stand above your arguments and elections." Our opinions are not in the strategy focus anymore and can be ignored without consequences. In the view of the WMF, the readers are not reading the articles because we editors have written them in a way that is well appreciated, such that it raises multi million dollars of donations a year. The readers visit them because of the great software that WMF develops. They would read anything, if it was just presented by their great software.“

Howwi ([8]):

„These ultra-soft-washed, excessively-complimentary periphrases of disregard leave me fearing that the respective self-understandings and goals have by now become incompatible. Irreversibly. To put it simply: „Here“, among the active contributors, we are mostly unpaid idealists, „there“ increasingly fulltime business-strategists. This will, sooner or later, clash... Or does this only seem inevitable to me?“

Aschmidt ([9]):

„They approach Wikipedia as if it was a company and we were their employees. [...] At the moment, this business is exploding with a speed even i would not have imagined...“

Hosse ([10]):

„after having read through the empty words by Lila and Erik, i am more and more coming to the opinion that this was it for me. In the mentoring program, i already stroke out myself, as i cannot be responsible for the exploitation of even more people.“

Rich ([11]):

„It is stunning that this happened. I see no respect for the community or consensus, let alone "love" or "kindness" here, this is the very behaviour that we relied on developers not doing... The probability of a serious fork grows by the day. We need to re-think the community-foundation interface... To some extent super-ing the German Wikipedia is worse than doing it to the English...“

I hope you can see the wider context your (i.e. WMF-staff’s) actions and their guiding strategy are seen within by this. Now would be the perfect time to show us what your vision really is - not merely by producing more clarifying circumscriptions for it, but by producing welcome outcome (regarding which, see the first part of this thread, as well as many other voices here and elsewhere).

Regards, Ca$e (talk) 09:56, 13 August 2014 (UTC)

I just want to clarify -- we are talking about software changes here, not content control, correct? Because it seems that people are really conflating the two concepts. LilaTretikov (talk) 17:23, 13 August 2014 (UTC)

Yes and no: We are not merely talking about "software changes", as if those were on a neutral ground and to your (i.e. WMF's) avail. We are talking about the ad hoc (mis)use of developers to establish a new hierarchy against locally elected maintainers and community members, and (mis)using this hiearchy to intervene against a binding consensus in a local community. Talk about "content control" is absolutely besides the matter, in this we seemingly agree. (I do not, however, see that anyone (anywhere of topical relevance) would "conflate the two concepts" - or if you will insist, maybe some WMF-staff-members did, however, when they tried to legitimate their measures as "office actions", as this concept is by definition only applicable where content control is of the essence. However, according to an overwhelming and well grounded perception, there was not, is not nor will ever be any legitimation at all for said measures.) On further points that may be of relevance regarding the actual topic at stake, please see (among others) my other posts on this very page. Ca$e (talk) 18:15, 13 August 2014 (UTC)

Well, the deciding point is: it are the editors of Wikipedia, mainly the big not bot-generated language versions whose work for Wikipedia is triggering the donated money the foundation is spending including the salaries you are paid from. So I understand your statement above from 21:00, 12 August 2014 as a spit in the face of every user ever improved an arcticle of Wikipedia. That's disappointing. Whatever the WMF does it cannot act against the consensus of the community. By the way, you should read and understand en:User:Jimbo Wales/Statement of principles, which is official since October 2001. Item four is the most important in this context, but point 7 is crucial as well. --Matthiasb (talk) 10:09, 13 August 2014 (UTC)

looking at jimbos principles Animal Farm comes to my mind. They had commandments, we have principles ...Sicherlich ^Post 10:25, 13 August 2014 (UTC)

Dear Lila, adding to ca$e's already rather extensive and well-written post, I would like to stress one point. You have talked a lot about the goal of providing free knowledge to the world. I hardly think that anyone here would deny that this is what we are doing and that this goal has a lot of appeal and has been responsible for attracting users: readers and in particular editors. However, the way that wikipedia works, i.e. that is a community-driven and self-governing (!) project, has played a probably equally important role in attracting editors and this sovereignty of the community has become very dear to us (at least on de.wikipedia, but I suppose this holds for other language versions as well). I feel that WMF either very much underestimates how much this means to the communities or simply does not care because its focus lies elsewhere (because you believe that the readers' interests are different and more important than the editors' or whatever other reason). And this perceived disrespect of the communities is what has angered lots of editors, as evidenced by the statements above both made here directly and quote by ca$e. I fear that wikipedia will lose many of its (best) editors (which are the foundation of wikipedia's success as others before me have pointed out), should the WMF's attitude towards them not change. --Null Drei Null (talk) 11:28, 13 August 2014 (UTC)

Thank you, Null Drei Null. I think you bring up and interesting point. We latch on to actions we take to achieve some basic goals, but we trigger bigger issues that may not even be clear are there. I have a very very deep respect for our editing community -- every time I meet editors I am amazed at the devotion and commitment there. By no means do I want to disrespect them. I don't think the issue here somehow is editorial sovereignty however. Even though from an neutral point of view we are simply talking about basic change management practices that are baseline for consistency and security of the site -- somehow it is perceived as infringing on sovereignty of editors. It seems this is not logical, it's emotional... Logically, to me it's a real surprise that people would try to change behavior of software for millions of users -- and at that, not improve, but rip-out components. Are we confusing "sovereignty with control? Somehow we need to figure out a method in which no-one is imposing, but we are using representative data (statistically relevant sample) to tell us what to do. -- LilaTretikov (talk) 17:23, 13 August 2014 (UTC)

Dear Lila, thank you for your answer, which makes very clear that there are misperceptions of essential parts of what is at stake here. We are not at all talking about "consistency and security of the site". We are not at all talking about "ripping-out components". (Everything that the component MediaViewer is technically capable of is available without it already, and, according to a large majority, currently in much better form, while many essential features are not provided anymore with MediaViwer in its current form.) Indeed, we are talking about "infringing on sovereignty of editors". The unprecedented establishment of a new user group, the provision of overriding user rights to a user group (= WMF staff) without any mandidate by the relevant community, and the overruling of elected local administrators in a matter where strategic issues are at stake held by the party (mis)using their already highly proplematic super-rights is nothing less than a most severe infringing. Indeed, there has never been a more grave infringement in the history of wikipedia as far as i know of. I strongly advise you to reconsider your perspective on this matter. Especially your talk of "emotionality" and "illogicality" is totally besides the issue and can be seen as quite provocative in the light of recent developments. Before "we" can talk on any new "methods", you should follow the steps explained e.g. by Rich and odder above. Ca$e (talk) 17:55, 13 August 2014 (UTC)

Thanks for clarifying. I leave all the administrative decisions now entirely to WMF and will invite a WMF employee for code review each time I file an edit request at Commons. I just wonder whether you have the resources. Good luck. -- Rillke (talk) 11:53, 13 August 2014 (UTC)

Just so that you, Lila, and others whom it may concern, know: Rillke is referring to his request for removal of adminship (which he had accepted to postpone until an official statement from your part would clarify things) and other rights, having to date been one of the most active and respected admins (among other very important work) ever. He is quite far from being the only one undertaking such considerations as an effect of this development. I strongly advise you and all others who have a say in the matter to read his general remarks (as linked to above) on the perceived direction WMF has taken. I am deeply sorry for his decision, but, as things seemingly stand, do understand and respect his reasoning. Ca$e (talk) 12:08, 13 August 2014 (UTC)

First, contrary to the comment above by Aschmidt, the editors are not your boss, we freely licence our work for others to change or discard, and most all of us do not do it to grab any power to be someone's boss. Second, it makes perfect sense to protect the software common from hacking, deletions, and edit wars. Some users appear to be confused that they are not freely licencing their work on a privately owned website but their confusion makes little sense, and is no basis for making any decisions. Alanscottwalker (talk) 12:11, 13 August 2014 (UTC)

Again: it is the editors' voluntary work which builds the content of WP, and it is the content of Wikipedia (which some call "free knowledge") which triggers donations. And it are donations who pay Ms. Tretikov. So, Alanscottwalker, who is the boss? Little hint: it is not the foundation. --Matthiasb (talk) 12:28, 13 August 2014 (UTC)

If you come here as an editor to be the boss of someone, these projects are not the place to do it -- pure and simple. Sure, it's editors freely licenced work (and, of course, some (whether I like it or not) get paid for what they do) that builds content, but most of us do not build content to cause people to donate money, nor to pay Ms. Tretikov, so your arguments are all irrelevant, as well as looking power hungry. Alanscottwalker (talk) 14:25, 13 August 2014 (UTC)

It has nothing to do with being boss, but with morals. The donors keep the Foundation and Systems running. The donors come in, because of teh content created by the editors and photographers and so on. Not because people think the presntation looks so neat, at least those are exceptions. So governing the communities with force maybe is possible, but it is horribly wrong by moral standards. And as a Foundation the WMF relies heavily on its moral good looks to people. Anyone organization who has similar goals like human rights, animal rights, spreading press freedom ... whatever relies on it's good reputation. By using force on your volunteers you are in danger to loose reputation. --Julius1990 (talk) 12:39, 13 August 2014 (UTC)

What's "immoral" is your claim that there is any extreme amount of force. That is just so baseless and untrue on multiple levels. As for the rest, see my comment above. Alanscottwalker (talk) 14:25, 13 August 2014 (UTC)

I introduced above the concept that rather than one party being the "boss" of the other, if we cannot simply be friends and colleagues, the employed volunteers (the Foundation) should consider the unpaid volunteers (the Communities) to be their customers. Alternative models are possible but this one is simple and well known - customer service is key. It might also be mentioned that any reader can join the community, there is no requirement to contribute content in order to contribute to discussions - we welcome ideas regardless of their origin. Rich Farmbrough 14:33 13 August 2014 (GMT).

Glad to see this

Hi Lila. I just wanted to say, I'm glad to read see you engage directly. Thank you for talking to the community directly, this small first step will go a long way in convincing some folks the top management doesn't live in ivory towers. Some of us might be less than cordial and rough around the edges but your patience above seems promising for the future. I would have wished you engaged on other venues as well but this a promising start. I'd also ask other community members to not monopolize the conversation here. Thank you. Theo10011 (talk) 08:20, 13 August 2014 (UTC)

P.S. sort of a minor tip, you might notice some of your replies don't have a date and time stamp, just your signature because you are probably using "~~~" instead of "~~~~" which leaves only your signature. Ideally it's always better to leave a date and time stamp with signature to distinguish between times and look at the discussion in context. :)

Thanks, Theo10011. Not sure I will be able to do as much of it all the time, but I think this is important. I'd like to see us figure out how to so this better in different languages, at scale. LilaTretikov (talk) 16:56, 13 August 2014 (UTC)

Timing

Yes, thank you thus far, but still I think your statements are moved by a wrong perception of what has actually happened. It seems to me that Erik and other German-speaking staff members have not translated the outcome of de:Wikipedia:Meinungsbilder/Medienbetrachter correctly. The poll was not a consequence of the community's general conservatism, or luddism, or whatever you seem to think. The German Wikipedia community does not oppose software improvements in general nor even the MediaViewer specifically. We don't think that the media sites on Commons are the best possible solution either (rather to the contrary, they have many problems, too). And finally, we do think about our readers, a lot, and do consider the (high) possibility that they like the new solution better. Still, what we (those who voted against its immediate implication) argued was that the feature has still too many problems to be implemented for everyone on an opt-out basis - for now. There are some serious issues concerning republication of images (compliance with licenses, attribution of authors/photographers) and the display of important media information within the new feature. What we said was simply: These issues need to be addressed first rather than meddled with hastily after going online for everyone. There are several counterarguments to this position, and, trust me, all of these did of course come up during the poll, as is characteristic of any Wikipedia discussion. However, these counterarguments did not gain a majority. Now, an unproblematic and (it seems to me) reasonable way to deal with the poll would have been to accept its outcome (and therewith the German community), to withdraw the new feature for a while from the German Wikipedia, to let your software staff work on it, thereby to consider the arguments brought forward, maybe (a pipe dream) engage in a discussion with the coummunity - and finally simply to reintroduce the feature without anyone complaining (well, there's always the odd complainer, but nevermind him. At least you would have had the majority of the community on your side).

You can still do this. --Tolanor (talk) 11:48, 13 August 2014 (UTC)

Your point, Tolanor, is basically the same I made here in German to Eric. --Matthiasb (talk) 13:18, 13 August 2014 (UTC)

It is a bit strange to see this being such a big deal given that the feature has been in Beta for nearly a year, was rolled out almost everywhere else in April with no issues, and has been on the de site since early June. So clearly it has not broken things. Why did it get so "hot" after two month of being in production, without reader complaints? Just wondering... LilaTretikov (talk) 16:56, 13 August 2014 (UTC)

Please see, among several others, my post above: The "hotness" did not at all occur because of the existence of MediaViewer, but regarding a) the evaluation of its status in its current form (including if it could be currently termed a "feature" at all!), b) the way it should in the current state be provided (as default? for which user group - visitors and logged in editors?) and above all c) the way both issues were, according to an overwhelming perception, mishandled by WMF-staff (on this, see my and several other posts above and especially this action by Fabrice Florin, which has been regarded as highly ignorant of a binding local consensus and thus was a main cause of said "hotness"; d) what's more, what in the last instance has created, according to lots and lots of community members, quite a few of which having already declared not to be willing to furthermore be part of such a climate of over-regulation by WMF, after tens and hundreds of thousands of edits and other contributions to wikipedia, wikicommons and in other areas, a totally inacceptable state of the matter: the introduction and (mis)use of a new "superprotection" level. On the steps you (i.e. WMF-staff/board including especially you in person) now should take, see especially the guidelines provided by Rich and odder. Ca$e (talk) 17:34, 13 August 2014 (UTC)

+1 to nearly all statements done by Ca$e especially 17:55, 13 August 2014 (UTC) Kein Einstein (talk) 21:21, 13 August 2014 (UTC)

Hi Lila, I cannot talk about how hot the topic of the Media Viewer was before last weekend. It was not hot for me, because I deactivated it from the first day. What wonders me is, what made this topic so hot for the WMF, that they first posted a statement, they will not respect the voting on the German Wikipedia (I would have even lived with this) and then escalated things on Sunday: conducting an Edit-War and finally inventing in a hurry a new Super-Power-Right to win the Edit-War. What made a topic like the enabling of a Media Player so hot and urgent, it could not be handled in a normal way by the German community and their local sysops? What made it so hot and urgent, that for this single problem, a user-right had to be invented, that had no guidance (as the legal office actions have) and has not until now? This is disturbing me very much. What can I expect in future from the Foundation? Will they now always act this way and trying to dominate everything with newly implanted technical features like Super-Protection instead of talking with the local communities and let them handle the situations their normal way? --Magiers (talk) 17:37, 13 August 2014 (UTC)

@LilaTretikov:, this is an excellent question, and one I hope the WMF staff is exploring in a diligent way. Truly, I think, the most important question is: how did the WMF staff involved fail to anticipate the strong negative reaction? Of course, it is difficult to predict things like this, but it's not impossible. In fact, I predicted it back in February; while I got a polite acknowledgment, it seems that the substance of my comment was ignored. I know of at least one other longtime community member who made that same prediction (in a private email to me, not long after that) -- and if I can get permission to share that email, I will do so.

So I think a very important corrolary to this question is: why were some Wikimedia community members able to see clearly that this software would spark a strong reaction, but the Wikimedia Foundation on the whole was unable to see it? -Pete F (talk) 18:01, 13 August 2014 (UTC)

I think the answer is illustrated rather well clearly by the use of the phrase "these people" in the ‎21:06, 21 August 2013 comment in this thread. There's a very clear us-and-them divide revealed in that little exchange. That sort of attitude towards the volunteer community is inevitably going to make it hard to hear unwelcome messages. Deltahedron (talk) 21:39, 13 August 2014 (UTC)

On the two months later, WHEN is ever an appropriate time? Currently, none. I think "after two months" is more reasonable than you make it sound, for two reasons a) the multimedia team is (hopefully and at last) switching to other tasks, i.e. the multimedia viewer should enter the so called "maintenance mode" and be wrapped up, hence it's reasonable for people to judge it as is; b) the long discussion and wait shows the community has been very careful and thoughtful in its decision making... after such a long discussion, the moment comes when you have to state a conclusion. --Nemo 09:40, 14 August 2014 (UTC)

Lila

Dear Lila Tretikov, we haven't met in London, but your keynote inspired me to a little parody. I prefer to post it here rather than that you spot it elsewhere on the web and wonder how the hell it might be meant. :-) Sometimes I feel that a twisted kind of poetry and humour helps me the best to express myself than other genres.

I don't believe in unicorns and mavericks - but I believe in a fresh perspective from the outside, evolution and occasionally hard but responsible decisions. As German songwriter Wolf Biermann said: Only he who changes remains himself. (Nur wer sich ändert, bleibt sich treu.)

Welcome, and kind regards, --Ziko (talk) 14:27, 13 August 2014 (UTC)

*The song parody was deleted with [12], Ziko was warned,[13] and there was a request for admin attention filed and closed.[14] --Abd (talk) 18:09, 13 August 2014 (UTC)

-- :) LilaTretikov (talk) 15:31, 13 August 2014 (UTC)

All Things Chapter

A section for Chapter related conversations.

„It was only 190”. A reminder as to what a Community actually is

Dear Lila and Erik,

I’m writing this to clarify why the “Superprotect-affair” is such an important matter for us Wikipedians, for the local communities, and probably for the future of Wikimedia in general. In many of the Wikimedia staff’s writing on the matter, I have found the argument that only 190 users in effect voted for the proposal that the MediaViewer be opted in rather than has to be opted out individually. Okay, let’s look at the numbers: 190 users pro, 72 against, 19 neutral. That makes some 300 users participating in the poll, as opposed to (as Erik pointed out) several thousands or even millions who are potentially affected by the decision. That does sound like a ridiculously small number.

And it is. The problem is: This is the community. Well, not exactly: The community of the German (as in German, the language) Wikipedia consists of maybe some 1000 people. These are the people who very regularly edit articles and take part in discussions and talks, and therefore know each other, at least by nick name. Of these 1000 people, maybe half or less participate in polls regularly. They elect sysops, vote for featured and good articles, discuss proposals for deletion. Since not every one of these some 500 users is interested in every poll, it’s fair to say that 300 is a pretty average number of participants in a poll.

I think this needs to be stressed: Only about 1000 people form the community of the German Wikipedia (the third largest one), and only about 300 to 500 take part in important decisions the community wants to make by poll. Wikipedia is a village. (Or maybe more like a Greek polis.) Of course, there are a lot of other people involved: Those who only write articles without taking part in discussions. Those who contribute one article and never return (like PR employees writing their company’s entry). Those who irregularly correct a spelling mistake when they see one. But all of this would not work without the community. These people do the basic work of Wikipedia: Wikify entries, check facts, delete vandalism, check the copyright status of media and texts, delete rubbish, and so on and so forth. Most of them also write a lot of the articles, mostly with a high quality status, some featured. Because featured articles and their like don’t stem from people who don’t feel committed to Wikipedia and care for it. You have to invest months of reading, thinking, and actual writing to complete only one featured article.

And this is the point: Even though what I call the community is not the whole of Wikipedia, it is these people who really care for it. It is these people who spend hours a day revising articles or discussing bread, or quarrelling over some stupid software feature. They only do this because they think it is important. And they are the only ones: The readers certainly don’t think some bread or, for that matter, some software feature is important. But the Wikipedians, the community, care on a whole other level: They spend hours of their day and have spent years of their life contributing to this odd online “encyclopedia”. They feel their life is somehow connected to it.

The Wikimedia Foundation often talks about the so-called Wikimedia or Wikipedia “movement”. That’s right: There is a movement. But even though Wikipedia is one of the most important sites on the internet with hundreds of millions of readers, the “movement” behind it consists only of these few people: the community. Each year, only a few hundred people attend the “WikiCon”, the largest meeting of German-speaking Wikipedians. Here you see them gathered again: The community, the village – the people who voted against the MediaViewer. These 300 people are the community. Wherever else you might look for it, you will not find it there.

Now, I realize that these people are, especially in a group (and they always come as a group), often very annoying. Online crowds look horrible from the outside (and not that much better from the inside). You can’t handle them, can’t organize them, and never get them to do what you want them to. But the beauty of Wikipedia is that you don’t have to. Wikipedia works without centralized governance. Communities have thus far been self-organized, and it worked. They voted their own sysops, made their own notability criteria, decided themselves whether they wanted their Wikipedia to opt in or opt out of an ill-conceived software feature.

But centralized, coercive governance is exactly what you want to introduce here. You don’t trust the communities. A Wikipedia poll always comes along with arguments and arguing, with mile-long texts written to convince others, with pro and con. And pro won this time. Now, you could have said to yourselves, “Well, they have decided. So what is the rationale behind this decision? What exactly do they want? What is their critique? How can we handle this?” Instead, you said, “What a bunch of idiots. Always against everything new. Conservative to the core. And it’s only 190 people anyway. Some ridiculous nerds not wanting us to meddle with their toy. We can’t give in to that, can we?”

The sad thing about the whole situation is that it is so wholly unnecessary. It would have been so easy to prevent it. Again: What we wanted was a better MediaViewer, not its complete abolishment. If you would have read what our actual arguments were, you would have seen that there’s no real problem. We did not say the Foundation was not allowed to introduce software changes. We did not say it wouldn’t maybe be better to have a new MediaViewer, better arranged than the old Commons sites. If you would’ve just listened and taken us seriously, none of this would have happened.

Now, at least, you should take the communities seriously. Even though what they’re doing looks quarrelsome, ridiculous, and messy, it works. What’s more: It would not work without them. So you better stop trying to coerce us with newly invented superpowers, and start talking to us again. Otherwise Wikipedia cannot work. It is much too complicated, messy, and great to be governed solely from a bureau in San Francisco. --Tolanor (talk) 18:10, 13 August 2014 (UTC)

+1 to Tolanor, and one thing I like to add about the "conservative" communities: Writing an encyclopedia (which is still the flagship of the Wikimedia projects) is a conservative task, and this literally: it is conserving the knowledge of the world. The people, that do this, are even so conservative, they go to libraries and search in books for trustworthy informations instead just googling around and copy and paste what they find. Yes, it's tempting, to make everything new, better, modern, but don't loose Your conservative editors in the process, because their quality work will not be replaceable so easily. --Magiers (talk) 18:42, 13 August 2014 (UTC)

+1 to the above. I have no personal preference for or against the media viewer. Overruling an RfC by office actions is limited to cases with certain prerequisites. Additionally, office actions don't cover the implementation of a 'superprotect', which as a feature is, in essence, unsocial. So far, no discussion to delay the introduction of the (faulty) media viewer was held. Even if the media viewer had been flawless, i find it problematic to overrule community decisions by office actions, even more so, if you take into account that the origin of this conflict was the introduction of a software feature. --Ghilt (talk) 19:27, 13 August 2014 (UTC)

+1--Cirdan (talk) 20:15, 13 August 2014 (UTC)

+1 --Hubertl (talk) 23:09, 18 August 2014 (UTC)

• Above is a cogent description of how wikis work from Tolanor. However, I'll take issue with the description of response as "what a bunch of idiots." Rather, I'd suggest, the response has failed, so far, to stand out and be distinct from that, and could look like that. I do not agree that the wiki could not be "governed" from a "bureau in San Francisco." It could. But it would not then be Wikipedia, it would be something else.

• It can take years of experience and study to understand how wikis work, and even more to understand how they don't work, and how to improve them.
• Wikis are not easily understood through standard business models, though they can certainly benefit from application of business concepts. In certain ways, this is new territory, though much of what happens on the wikis can be understood from accumulated business wisdom and what is known about consensus process. Maybe it's time to get smart. --Abd (talk) 19:35, 13 August 2014 (UTC)

• You should accept the communities seriously. Overruling = lose*lose. --Gruß Tom (talk) 20:48, 13 August 2014 (UTC)

Some advice from an old Wikipedia hand

Hi Lila. It heartens me to see you are engaging with the community on your talk page. I only hope that doing this doesn't burn you out.

But to the point of my message. I've been a contributor to Wikipedia for many years -- I made my first edits back in October 2001, when there were a many, many more missing articles than useful ones. Based on this experience, I'd like to offer two pieces of advice. They will definitely apply to the English Wikipedia, but I suspect they may also apply to most of the other Wikimedia projects:

1. Although Wikipedia is a bottom-up run project, for some reason people forget that & repeatedly try to force it to do something -- which never works. Wikipedia is not an organization with rewards & penalties like a business or the military. At best, anyone who tries to force the community to do something fails; at worst, that person ends up either banned from the community, or leaves embittered & with a damaged reputation. Appeals to authority or coercive actions do not work in the long run, even if that person has wide support; eventually that person offends enough people & their birds come home to roost. I could cite many examples, but will limit myself to one prominent one: Jimmy Wales tried that with Pending Changes years ago, & the result was that it was not adopted, despite having wide support. AFAICR, instead of engaging who disagreed with him over Pending Changes -- or finding a clear & objective way to show that it would solve the problem with biographical articles on living people -- he tried things such as appealing to his authority, or gaming with the rules (when the RfC to enable it failed by missing the 75% approval bar by a few points, he suggested that bar should be lowered -- which only hardened resistance to the idea). I see the same thing playing out with this squabble over Media Viewer.
2. What works is talking to people. In my interactions with other volunteers over the years one theme is very common: most contributors feel their efforts slighted. Those who work on content are perhaps the most vocal about this, but I've seen this expressed by those who handle maintenance tasks on the Wikis (e.g. closing AfD discussions, clearing out backlogs, etc.), or maintain the code. Talking & listening to people -- finding out what helps them, showing their appreciation for handling difficult tasks (I can attest researching & creating content is far more difficult with certain topics than others, & is one reason for the systemic bias seen on Wikipedia) -- this helps keeping good people working on the Wiki. Doing this requires a lot of work, but that is what effects lasting change at Wikipedia. And I don't see this being done on a significant basis: for example, I asked twice what the point of Media Viewer was -- all successful software must be designed to address a defined issue -- & no one ever bothered to tell me. Or point me to its description, if one exists. Instead, community outreach people regularly appear to be in permanent damage control whenever they post at the Village Pump of the English Wikipedia: instead of reaching out & building support for new ideas & possibilities, they are reacting to angry volunteers who believe they are struggling with Yet Another Example of their contributions being slighted.

I know I don't always explain my ideas clearly & fluently (& pardon my jargon, I've been an Internet person since my days years ago when I worked phone support for Netscape 1.1, so sometimes I can't think of a better way to express myself) but I hope my two points above make some sense. -- Llywrch (talk) 20:13, 13 August 2014 (UTC)

Llywrch -- thank you for the kind words and trying to help me out here. 1. I am puzzled about why do people feel slighted -- where is it stemming from -- it definitely seems like a long standing sentiment. 2. How do we have a process that is sensical around software roll-outs that also fits with our ethos. I agree we need to listen better and somehow reach our users better as well. But we need a way to get there from here. MV is just an example. I don't know why that feature got prioritized specifically, but I can tell you that our user research shows that users (not editors, because editors are already really used to the current UI) expect a certain behavior from on-site interactions. Specifically they expect a zoom-in image/lightbox when they click on a thumbnail in the article. This is a mental model most users have because there are patterns on the web that are common. So it makes sense for us to implement something like MV (probably not as the top priority, but we are here now). So let's say we are doing it -- we put up a note for people to engage last November -- it gets removed. We do not force it to stay. So now, people are saying they did not know and did not engage until it is rolled out to over 800 wikis -- all but 2 with no pushback.... Now that we are here... Why not do an RfC on how to improve? Why force it out to where it is impossible to actually make better... I don't know how to make sense of it. -- LilaTretikov (talk) 22:17, 13 August 2014 (UTC)

Dear Lila, you wrote I don't know how to make sense of it. As you are businessexperienced you'll perhaps understand this: "Never Argue with a Customer". WMF should appologize concerning Eriks action. Common treatment of cases in WP is reset to the state before struggles and find a consensus solution. --Gruß Tom (talk) 22:46, 13 August 2014 (UTC)

"I don't know" is the most hope-inspiring comment I've seen in this affair. It is when we think we know that we get into trouble! The WMF is not a business with customers, it is a nonprofit with a community of interest, and such frequently develop some split or gap between the community and staff. Even within the community there can be a gap between the larger community and the most active core.

The staff form a coherent group with clear responsibility and cannot be equated with the community, which often is lacking structure (it's common for a community to form an organization and think that the organization is its structure, and then, as staff is hired and conflicts of interest arise, the community has no independent structure.)

If staff recognizes that finding genuine consensus with the community is of very high value, and acts to facilitate such consensus, it's win-win. Good luck. --Abd (talk) 23:46, 13 August 2014 (UTC)

Ah, by no means do I have all the answers, but I have a lot of questions. All I know is the only "cardinal sin" in product design is indifference towards users -- I hope we agree there :) We need to agree on how, thought. Because we have so many, and so different... -- LilaTretikov (talk) 00:40, 14 August 2014 (UTC)

But this is rather an "I don't know" as in "I don't know, and I don't want to know, it just seems stupid". We can explain to you the severe problems we have with MV. We did already, in the poll. Erik actually knows about the problems, he expressed them here. He can explain them to you, too. In our estimation (this may also have to do with the strict understanding the German community has concerning copyrights) it is a major flaw that in the new tool author's names are only showed "if machine-readable". It is also a major issue if image information is only given "if machine-readable". You don't seem to see this as major problems, but we do. And what I and a lot of others just want you to understand is that we have good reasons to do so. We are not stupid, completely irrational, or whatever you might think. One of the lines of argument, in short: The licenses under which we put our media contributions – such as CC-by-SA – state that our names have to be credited. This is legally binding. But now, the new tool often doesn't even show our names. How is anyone supposed to reuse our media correctly, stating our name, when not even our own MediaViewer does it? Some might say that the MediaViewer, as it is, results in fact in a mass violation of copyrights. You might have a different opinion, but we simply ask to have our opinion counted, and to be able to express it within our own Wikipedia, which we have built and up to now governed ourselves. Normally, in Wikipedia, if you want your opinion to be accepted or even become "consensus", you need to convince the others with arguments. You, instead, used force and coercion. --Tolanor (talk) 15:45, 14 August 2014 (UTC)

"This is legally binding. But now, the new tool often doesn't " - if it is so: maybe an accusation against the WMF would help? There are millions of pictures on the Server of the WMF; for sure a lot by german fotographers. ...Sicherlich ^Post 15:50, 14 August 2014 (UTC)

Just had the same thought, maybe we should explore this further. See de:Wikipedia:Urheberrechtsfragen#Massenhafte URVs durch MediaViewer?. --Tolanor (talk) 16:06, 14 August 2014 (UTC)

It is good that you have articulated the objection that is not just "I like the old way". Nonetheless, as would be expected it is a legal objection. Legality of what the WMF does is and always has been up to the WMF (or a court) and not a communal decision. Alanscottwalker (talk) 17:04, 14 August 2014 (UTC)

"I am puzzled about why do people feel slighted -- where is it stemming from -- it definitely seems like a long standing sentiment." - Thank you, i completely agree with Abd on this reply giving us some hope that, after this issue will be resolved, not all may be lost regarding the for some time quite tenseful relationship between WMF and the most active parts of grounding communities -- communities it should help and not, repeatedly, slight. Perhaps you have not been aware that this situation is only the currently last one in a series of misbehavior on WMF's part, one of which concerned image filtering capabilities, for example. What we have seen in this series has been comparable to the current situation in that there has been a desastrous way of communication and abuse of capabilities on WMF's part, it getting worse and worse especially over the last about 3 years, now having reached the downpoint, at least until your direct involvement in the communication. Not ever, at least to my knowledge, has WMF previously escalated things that drastically by one-sidedly establishing a new user hieararchy, by threatening admins with revoking their rights (also something which has no base at all in the guiding principles at work here!) and (mis)using this new hierarchy to directly overrule in local contexts. This misbehavior is perceived in a series of conflicts caused by WMF and its stance towards strategic implementations of technical "features" (like image filtering) and surrounding communication. Exactly the misuse of developer resources and technical features to overrule local consensuses has been what has multiple times caused well-founded uproar in communities. Exactly that is what has been done here, only to a larger scale, as fundamental principles are concerned when it comes to who regulates user rights and decisions for whom. To me, as to several others, it is absolutely no wonder that many people feel slighted (to say the least) -- instead, it is exactly what we could have foretold you, or to be more precise, did foretell you, and especially in regard to those users who are by far the most active and engaged. If you do not succeed in revoking and resolving the current situation, in making amends and trying to win back users that have been a longstanding basis for their respective communities, you, or your colleagues, will have already produced more damage than you will ever in, say, 5 years at the minimum, be able to repair. This is nothing you can solve with money, with PR-speeches, with advertising or hired manpower. What is needed is a complete change of perspective and behavior on WMF's part, with corresponding, visible and felt outcomes. To repeat, the current situation is only the last one in a longer series -- a series which many active community member remember well; the latest escalations on WMF's part are therefore seen as proof that as WMF is concerned, the fears have been well founded. Ca$e (talk) 07:15, 14 August 2014 (UTC) PS: On the same topic, see also this post by Henning. Ca$e (talk) 12:27, 14 August 2014 (UTC)

Lila, the short answer to why contributors feel slighted is that, even under the best circumstances, communication is difficult, & passive communication -- such as posting notices -- often fails to get out the word. Most contributors simply want to contribute, to make the content better. However, when a wide number stop their editing to say there is a problem, their concern needs to be taken seriously. In this case hundreds of Wikipedians from different projects spoke up with their concerns. And the response was to claim that thousands of otherwise unknown & anonymous users thought differently -- a response that evokes the old Usenet response, "the lurkers support me in email", often invoked by the party who has lost the argument but refuses to admit they're wrong. Your response above was far better, because it gives someone like me something tangible to respond to. I could ask for a link to that report, & see if its methodology is better than our own experience. I could see if the number sampled is comparable to the number of Wikipedians who protested this change. And I could see if the persons who produced this study did not make a logical fallacy, that one choice must be better than the other, instead of accepting different choices can be equally good. By this I am thinking of the case where a Mac one-button mouse is different than a three-button Windows mouse, yet both are equally intuitive. Or the case where making a windows active in MS Windows is different from the procedure in XWindow -- clicking on the window vs. moving the cursor over the window.

I could say a lot more about communicating with the community, about how & why we feel we are isolated & disenfranchised, but the best thing for you to do would be to actually go out to the project Wikis & see what happens there & how we work for yourself. (I believe one cause for the disconnect between the volunteers & the Foundation staff is that they've forgotten how Wikipedia actually works. And why unilaterally forcing any project to do something fails.)

As for the Foundation's process in rolling out software, as a professional QA engineer I find myself agreeing with Kww's comments below, that the process as it works now is misdirected & broken. Please read & consider his words there carefully. -- Llywrch (talk) 20:10, 14 August 2014 (UTC)

Hi Lila, thank you for taking the time. Firstly, many of those who voted for an opt-in for MV did not reject this project for all times but were of the opinion that its introduction was premature as there were a number of serious problems with it. Among the serious problems are misrepresentations of the license and/or the attribution (credit line) for some non-trivial subset of Commons. (Examples for this can be found in the discussions.) This can be solved but it should be obvious that an opt-out deployment of this feature is premature as long this is not settled. The deployment of new features in alpha state is a repeating pattern where many of us would like to see a change. Secondly, the expectations and mental models are different in Wikipedia. Besides license templates, geolocations, EXIF data etc. there exist in many cases interesting content in the image descriptions which can be edited. This is effectively hidden by the MV as many readers are unlikely to find the Commons page behind the MV page. Take for example this photograph which is used on multiple Wikipedia projects. In summary, I would like to second Tolanor's comment above: Please take the voices of the poll seriously. Many of the 300 who participated in the poll have years of experience with the projects, are admins at de-wp or at Commons, know how media at Commons get reused in practice, have worked with external partners and readers etc. I think it is safe to say that many of them do not speak just for themselves in their personal interest but from their individual perspectives and experience in the interest of the projects and our overall mission. Regards, AFBorchert (talk) 23:28, 13 August 2014 (UTC)

The core problem

The core problem is that the software staff is using inapplicable models to compare Wikipedia to, makes bad decisions as a result, and then they don't tend to listen when we tell them what a bad decision they have made. Media Viewer is a minor nuisance compared to the the disaster that Visual Editor was: when I turned Visual Editor back into opt-in and took it away from unregistered users on English Wikipedia, I had near universal support. Erik Möller's and James Forrester's response was to finally back down, but not before they spread falsehoods about the nature of my fix and tried to pretend that they had saved English Wikipedia from a bad patch. In fact, my patch did precisely what it was intended to do and did it well. VE was worse the MV because it actively damaged articles and required constant vigilance by all editors to contain the damage. MV just misses the point: the purpose of an image in an article is to illustrate a point in the surrounding text. It's not intended to be viewed out of context, and there's no reason to expect that anyone would want to page through the images in an article (or, if they do, they are an unintended audience: someone that isn't interested in reading an encyclopedia article but just wants to look at the pretty pictures). We now have Flow coming along, where Maryam has made it abundantly clear that one of her purposes is to disrupt our communication patterns with techniques ranging from intentionally placing topics in the opposite order from normal to limiting the depth of our branching to preventing us from being able to modify comments. At least part of the reason I object to the new "superprotect" state is my expectation that when WMF attempts to make Flow universal, superprotect will be used to ensure that we have no choice, no matter how problematic the software is.

All of these things seem to stem from forgetting what Wikipedia is: it's a user-written encyclopedia. What we need software to help with is making certain that we can properly cite sources, that we can properly find sources, and that we have places to archive sources so that they stay stable. We need tools that focus on allowing us to properly cross-reference information between articles, automatically find places that articles contradict each other so that we can resolve the contradictions, and allow us to bundle information together to make it easier to reuse between articles. We need tools that will allow us to more easily identify disruptive editors and prevent them from editing.

Instead, we get software that makes it easier to chat. We get picture browsers. Because the designers seem to compare us to things like Wordpress, Facebook, and bulletin-board sites, we get editors that don't solve the problems we have creating articles, but actually couldn't edit simple articles without damaging them. Part of that problem is because the WMF believes in the myth that releasing broken software to get user feedback is a better approach then understanding what the software is supposed to do before starting to code (and yes, I'm a computer professional, working in communications and avionics). Even people that believe in Agile will tell you that what the WMF practices isn't correct Agile development.

And why do we get all these strangely broken and inappropriate tools? Apparently because people worry about how many people don't like to work on Wikipedia. That's not surprising: encyclopedia writing is a strange hobby. Not many people have the temperament for it, and even fewer have the skills to do it well. I've seen statements decrying our low editor retention rate, but I think those miss the point: 1% is actually remarkably high. About 0.06% of American collect model trains. Somewhere around 1% of the world's population collects stamps. Coin collecting, Pokemon battle cards, you name it: I think you would be hard put to find a hobby that goes above single digit percentages for a retention rate.

I've both worked for and ran companies, and never have I seen such an obsession with how many workers we could accumulate. Normally, people focus on the product. Are we falling behind? Does the world need substantially more than the 4.5 million articles we have on hand? I think not. I would like to see WMF focus on the tools we need to improve the product.

I would also love to see a moratorium on lying. A commitment that the next time we can point out that Phillipe, Erik, James, or any member of the WMF staff is telling bald-faced lies, that staffer would be immediately terminated without hesitation. That commitment alone would do wonders for interaction between the WMF and the various communities.Kww (talk) 06:21, 14 August 2014 (UTC)

thanks, i couldn't agree more. Ca$e (talk) 07:20, 14 August 2014 (UTC)

Well, circumventing Flow will be very easy: the community will move away from talk pages entirely. Indeed, already now, most discussions take place on non-talk pages, pages like Wikimedia Foundation Board noticeboard for example. There is absolutely no way to enforce flow, and already now it can predicted that Flow will just be yet another software feature disaster. The problem starts that obviously the Flow devs never saw talk pages like en:Talk:Ferguson, Missouri. Flow will break the functionality of English Wikipedia's quality assessment system (see also en:Wikipedia:Version 1.0 Editorial Team/Assessment) – note the bunch of categories the talk page is sorted in, and non of them has to do with talking (there are just about 4,580,000 articles with talk pages like that in EN:WP only), and how WikiProjects and/or portals in several language versions are organising their work (e.g. the dead weblink project in DE, see de:Kategorie:Wikipedia:Defekter Weblink Bot). Again well experienced users from different communities warn, that for good reasons they don't believe that Flow ever can be used sensefully, and again totally blind of reality the work on Flow goes on unpertubed. Flow is just another feature unasked for and unneeded and so far just yet another communication fiasko. --Matthiasb (talk) 08:23, 14 August 2014 (UTC)

I think it's implausible that being a Wikipedia editor is a suitable hobby for just a few thousand people per project. Why shouldn't we try to double or triple that number by providing a user friendly environment with uncluttered source text (VE) and structured talk pages (Flow)? --Kurt Jansson (talk) 08:48, 14 August 2014 (UTC)

//edit c.// Thx to Kww for his statement; and thx for the judgement of Flow by Matthias above. That is. -jkbx- (an internetcafe + vacation sock of -jkb-) 08:51, 14 August 2014 (UTC)

Kurt, Flow is not, in effect, about "structured talk pages", but about hopelessly broken talk pages. (As has been pointed out in detail more than a hundred times already and, as it is off topic in the current context, should not be repeated here. Very many people will never, ever, accept anything as from the ground up hopelessly broken and desastrous as Flow. If WMF wants to double the number of active contributors - which seems quite ridiculous, as they are currently and for several years doing their probable best in chasing away the current userbase from its core -, then it is speeding towards the totally wrong direction, as, e.g., Pete, and many others, have pointed out with an astonishing degree of placidity. There are many hundreds of things WMF could do that are really needed. Not too mention hundreds of critical bugs and minor feature requests way more important than MV, just one example that has been mentioned already on this page are improvements in citing and referencing - and everyone who would need VE to participate at all would need those anyway, plus, quite many who do not need VE nor want VE ask for those, too. WMF's investment of resources is totally of the line regarding what people actually need and want. But all that is rather off topic here, plus, it all has been well documented elsewhere.) Ca$e (talk) 09:15, 14 August 2014 (UTC)

Kurt Jansson, there would have been nothing wrong at all with an effort to provide a simpler method of creating Wikipedia articles: something that understood how templates and table syntax and references and categories and raw wikitext and html tags all came together to produce an article. Something that assisted a novice editor with all the complexity would have been wonderful. Instead, the WMF produced a tool that could handle something on the level of the average blog post, tried to block editing of anything it didn't understand, and frequently mutilated the pieces it couldn't understand. Here we are, over a year later, and it still can't edit tables, number references properly, or edit the contents of many of the more complex templates. When it was released, it couldn't handle editing all components of any article above the stub level, and it still can't handle editing all components of moderately complex articles. The constant refrain from developers was that our articles were too complex in terms of their formatting, which isn't even a debating position: whatever complexity they are, being able to edit all parts of all of them without corrupting them was the minimum standard for release. I don't think anyone even tried to achieve that. As for structured discussion pages: fine. Just make sure that the new tool is capable of everything our existing tools are before replacing them, and remember that "you shouldn't want to do that" isn't a valid opinion for a software developer when it comes to existing features. Kww (talk) 02:36, 15 August 2014 (UTC)

You're absolutely right that the VE had too many serious bugs when it was deployed a year ago. But the developers went out of their way to make VE compatible with even the most absurd behavior of the old parser. There are still bugs and there will be bugs in the future, like with every complex piece of code. But the talks the developers gave at Wikimania impressed me, I hope there will be videos for "VisualEditor — engineering against the odds" and "VisualEditor — helping users edit more easily" soon. And I really want the new references feature, it will have a very positive impact. --Kurt Jansson (talk) 12:15, 15 August 2014 (UTC)

Any presentation that includes a statement like "A year ago, we released VisualEditor for normal users to try out for real, giving them a tool that was basic and missing some common abilities, but already better for new users than wikitext." has a fundamental disconnect with reality. The problem is that VE attempts to hide the underlying structure of the article from the user as opposed to helping the user understand, manipulate, and control that structure.Kww (talk) 17:16, 15 August 2014 (UTC)

"I would also love to see a moratorium on lying." This would fix everything. A firing of staff members, a loss of privilege for arbcom members, and a temporary suspension for admin. Lying is the number one problem that seeks to destroy this project, and the vast majority of problems come from people abusing their authority at the top then lying as part of an attempt to destroy those who point it out. Ottava Rima (talk) 15:54, 14 August 2014 (UTC)

I love the idea of "a moratorium on lying." But I would have expected Wiki-experience to have informed us that just because someone is wrong, even demonstrably wrong, it does not follow that they are "lying" - indeed that is one of the reasons we have AGF. But apart form that (and some strange statistics on train-collecting) KWW's comment does seem spot on. Rich Farmbrough 16:27 14 August 2014 (GMT).

AGF is only in effect when a powerful person has been caught red handed and is trying to justify their abuse. AGF never applies to the people who do all the work but have none of the power. Standards must be applied in proportion to the amount of power an individual has. Ottava Rima (talk) 19:14, 14 August 2014 (UTC)

But the sad truth is that most of the human race will self-justify, and believe what they are saying. One purpose of AGF is to allow us to use platonic questioning to dispose of this false reasoning, and bring about a discussion that is based on facts rather than "You're a booby" - "Booby yerself!". AGF is not a suicide pact, and, to a certain extent we work with the well known proverb "By their fruits shall ye know them." Currently those fruits are discord (and crap software), that is what needs to be resolved, not a bad choice of epistemologically challenged argument or data. Rich Farmbrough 18:00 15 August 2014 (GMT).

A lie is based on what you say vs reality, not what you say vs what you believe. You can "believe your own lie" - the term is delusional. Ottava Rima (talk) 20:56, 15 August 2014 (UTC)

Sorry, and also for posting offtopic, but we use such expressions differently. See, e.g. [15], [16]. Ca$e (talk) 07:47, 16 August 2014 (UTC)

Is the WMF a software company?

Reviewing the Mission statement.

The mission of the Wikimedia Foundation is to empower and engage people around the world to collect and develop educational content ... and to disseminate it effectively and globally.

From Values

Our community is our biggest asset

We are a community-based organization. We must operate with a mix of staff members, and of volunteers, working together to achieve our mission.

We support community-led collaborative projects, and must respect the work and the ideas of our community. We must listen and take into account our communities in any decisions taken to achieve our mission.

Software is subsidiary to the primary goal, which is explicitly to "empower and engage people." That's the "community." There is nothing in the goals that is about readers, except, obviously, educational content is to be used somehow, in the end, "disseminated." The primary purpose of the Foundation and the software is to empower the collection and development of educational content, which is why "wiki."

In an eagerness to "improve reader experience," as assessed by software developers, core mission and core values, particularly the value of "respecting the work and the ideas of our communities" and taking them into account, may have been neglected. All those who develop content, the "editor community," are constantly reading the content, and are regularly engaged with it, more intimately than software developers.

People who develop software may become attached to it, it's easy to understand. It's their baby, after all. They put long hours into it. However, the very purpose of the Foundation and the software is empowering that community, and attempting to dominate and control the community -- and certain recent statements and actions have that appearance -- is backwards.

It is legally necessary for the Foundation to be independent, it cannot be "run by the community." Having said that, the Foundation's role is not to "lead the community," as if the Foundation were a government for the community, but to empower' the community.

I recommend that recent statements by Foundation staff that appear to depart from the core WMF mission and values be carefully reviewed and corrected by WMF leadership to make clear, now and for the future, the relationship of the WMF and the community.

Long-term, structures whereby the community coherently and efficiently advises the WMF should be developed. It is not clear that existing structures truly represent the overall community that creates the projects. Supermajority election, often used on the wikis, is known to generate warped representation, and may not be appropriate, at all, to create what would be needed, proportional representation. There are simple techniques, which would adapt well to wiki conditions, that do not require factions or parties. There are many problems in attempting to create governing structures this way, but creating advisory structures would be quite safe. --Abd (talk) 17:08, 14 August 2014 (UTC)

Dear Abd, maybe you have missed this. Evidently, we can give up on WMF being able to have its failures "reviewed and corrected" by itself. WMF-employees have crossed multiple lines, WMF-Board has explicitly supported recent aberrations, so good riddance to those values and former "community"! Ca$e (talk) 17:21, 14 August 2014 (UTC)

Lila is travelling. There is no statement from the WMF Board there. This whole thing blew up quickly. Give Lila time, please. --Abd (talk) 17:57, 14 August 2014 (UTC)

There is a statemet of the Board already. And i actually can hardly fit it to the mission statemt, the values or even Jimbos principles ... --Julius1990 (talk) 18:17, 14 August 2014 (UTC)

Julius is correct, that there is now a statement that is titled "Board statement on the Media Viewer roll out." However, it begins:

Some of you have asked the Board and its individual members for feedback. Some of us are already in conversation with you or are planning to answer on different pages. This is our general common statement:

And then it goes on. It is signed "On behalf of the Wikimedia Board of Trustees." That certainly looks like a "Board statement." However, this situation blew up over only a few days. It takes two day's notice to call a special meeting of the Board, and we do not know if such a meeting has been called. The email simply implies that "some" of the Board is "already in conversation," which actually looks like there has not been a meeting and a formal decision upon vote. Rather, there is a train wreck on all sides. We expect wikis to generate train wrecks. We do not so much expect it from professionals, but there have long been concerns about the professionalism of WMF staff.

I am expecting that when Lila is back in the harness, we will see some action to address all this, and we will ultimately see an actual Board resolution. I'm expecting nothing less. --Abd (talk) 01:32, 15 August 2014 (UTC)

The Board met last week, on 6-7 August, before Wikimana, as that statement says. The minutes have not yet been posted. Deltahedron (talk) 06:30, 15 August 2014 (UTC)

Right. That was before this blew up. It's obvious the WMF determined a course of action, but did not necessarily anticipate the depth and intensity of community response. So what we are seeing is simply a manifestation of that determination, inertia, not a rejection of the community response, which has not been formally considered. --Abd (talk) 15:46, 15 August 2014 (UTC)

First, developing freely licensed software is developing free educational knowledge, it is also the means of free distribution globally. If someone wants to work on a project that runs on commercial software, these are not those projects. Alanscottwalker (talk) 11:12, 15 August 2014 (UTC)

If you think these projects run solely on free (as in beer) or free (as in speech) software, you are mistaken. A huge amount of money has been spent on developing MeiaWiki software (Beer), and all the machines run non-free BIOS and microcode(speech and beer). I doubt the switching fabric is free, and certainly the delivery is across non-free routers, to largely non-free desktops and partially non-free handhelds running a both free and non-free browsers.

As to the code constituting content in its own right, this would be an extremely cost-ineffective way of generating content.

Rich Farmbrough 14:11 16 August 2014 (GMT).

Free, as in freely licensed, free, as in not charged for by takers; as for the rest sure, why not support giving away devices, commercial software, and connections to the needy or to the public access points, but that is rather beyond this organization (although it may do some of that, too). Alanscottwalker (talk) 15:06, 16 August 2014 (UTC)

Yet another opinion on what really is the problem

Dear Lila,

First off, welcome to your baptism of fire. Good to see you here. Best of luck as ED.

Now to the heart of the matter. The WMF has decided to overrule the community. Reasonable questions can be asked, and discussions can be had on whether the WMF has the moral right or even obligation to do that (and strong cases can be made on either side). This is nothing new, but a culmination of a process that has been going on for a few years now (AFTv5, VE, MV, soon Flow - where I don't expect anything else to happen than is now happening). Mutual distrust and even disgust has been growing over the years. What has happened now is a demarcation line, and if the rift is not glued a point of no return. Maybe that's OK; I believe in a different model, but what any one individual believes in isn't all that important. The important part here is not who blinks first, or who bows to whom. The important point is that when the WMF decides it can overrule the community it declares once and for all that the WMF isn't part of the community, but separate from it. That maybe the WMF and the community can work together and achieve great things together, but as separate entities and identities. That realization is very hurtful to me. Up to this point we were a community where the members employed by the WMF formed a group that had a strained relationship with the rest of the community. Now it has indicated it's not part of it anymore. It has forced me to stop thinking of the WMF employees as my fellow wikimedians who are on the WMF payroll, but as part of an organization with the same goals (and in some cases also part of some Wikimedia community, but no longer inherently). It seems you are OK with that, seeing your remark

"Accountability: As the Executive Director I am the person ultimately responsible for the direction and actions of the WMF. WMF makes all decisions as a team with my oversight, and we stand by these decisions together."

The team used to be all of us. Now it's you and the people on the payroll of the organisation you're the executive director of. You've taken away ability of people who work for the WMF to as part of that work be Wikimedia community members, because they can no longer openly have dissenting opinions that some decision is wrong, and not support it. There is no way these two can be reconciled.

That doesn't have to be a bad thing - I think it is, but maybe I'm being conservative and sentimental - this may well be a pivotal moment for the wiki movement, and it is your prerogative and responsibility as ED of the WMF to set the course of the WMF through this. Maybe a hard break is just what is needed. I want to be sure you know it though; this is not a future course that should be taken without realizing how much of a break it is.

Regards, a member of the community the WMF used to be part of, Martijn Hoekstra (talk) 08:55, 14 August 2014 (UTC)

Supporting one's staff is admirable. Supporting their wrong decisions is not. Rich Farmbrough 16:30 14 August 2014 (GMT).

Supporting staff is essential to good management. However, it is also essential to guide staff. The ED indeed, given the way the WMF is structured, is responsible, and it's clear that recent decisions had the approval of the ED. So, now, we, as the community, must ourselves, as we have the will and insight, support the WMF, which works for us, it is in the mission statement. If a staff member forgets the corporate mission and falls into some other agenda -- which might be a noble agenda! -- it's the ED's responsibility to remind him or her, and if the ED misses the point, it is our duty to remind her. And that's what we've been doing, only we are a community, not a coherent organization, and we exist in many different levels of sophistication and understanding, so, shall we say, some of our reminding has been unskillful. On behalf of the community, I apologize for that. In the future, as we develop skill and structure, we will do better. I promise. --Abd (talk) 17:15, 14 August 2014 (UTC)

Proper identification of WMF user accounts

This should actually be some basic organization: all Wiki user accounts of staff members should be assigned a (WMF) suffix for a clear identification. The highly controversial use of the self-made superprotect right was done by a user Eloquence (AKA Erik Möller) whithout a clear indication this was a WMF action. See [17] for a list, multiple users look like regular editors. In no way do I support these WMF actions recently done (showing a disrespect for local communities + threating local admins), this feels like a dictatorship. --Denniss (talk) 10:22, 14 August 2014 (UTC)

There is actually an RFC about this already here on meta: Requests for comment/Distinguishing Wikimedia Foundation staff accounts for official actions and personal use Zellfaze (talk) 13:18, 14 August 2014 (UTC)

Thanks for the info. I'm still surprised that this wasn't done from the get-go. --Denniss (talk) 13:56, 14 August 2014 (UTC)

Actually it would already help if WMF sticks to what, to my understanding, they wrote by themself as rules: en:Wikipedia:Office actions. The use of the superprotect-right was not because of "questionable or illegal Wikimedia content" nor was ist "by formal complaint made off-wiki " - it's just some random "I have the power, so I can use it". Who need rules. ... So please Lila just for the illusion; update en:Wikipedia:Office actions. Maybe shorten the whole thing to "WMF has the power, so WMF can use it whenever it thinks it is needed" ...Sicherlich ^Post 15:29, 14 August 2014 (UTC)

The official policies are on Meta, not on en-wiki, it’s this one: Office actions. And there doesn’t have to be written down any dictatorshop rules, just because someone doesn’t stick to the own policy. Not the policy should be changed, because it was broken, the people shall respect the own policy, that’s all. Especially those people who do office actions according to the policy: Office actions#Who does office actions:

"Certain other staff members of the WMF Office staff, such as the Executive or Deputy Directors, or the General Counsel."

The actions taken have been done as "WMF action" and "WMF edit". So I would like to have someone of the Legal Team to clarify, if this one about the Superprotect has been such an office action or not. It has already been taken as an example for such an office action because of legal reasons. But I don’t see no legal reason at all for this superprotecting the Common.js page. It would be good, if this would be clear. If there are any legal issues with the Common.js page, then please tell us which ones and if there are any complaints about it made off-wiki which I don’t see at all. I’m a bit confused, if this "WMF edit" made by a Deputy Director who has the right to make office actions according to the policy, is such an office action or not. Also at en.wikipedia, there have been discussions about it, if the action taken there falls under this policy or not, because it has been made as official "WMF action". --Winternacht (talk) 11:46, 16 August 2014 (UTC)

legal issues -- quite the opposite: by enforcing MV, WMF brought itself in the position where it knowingly (as it was pointed out many times, e.g. in May and also preliminary to the binding consensus in the german-speaking WP) invites breaching correct licensing in quite a lot of cases, to say the least, and notably in a global context. WMF once was part of a community movement about free information, now it seems to be a business willingly and knowingly engaging in affronting its former communities and breaking not only its very own guiding principles (as established above e.g. by Matthiasb) but also, to say the least, inviting breaches of copyright law in a wantonly negligent stance. Good luck with that! Ca$e (talk) 12:13, 16 August 2014 (UTC)

Yes, I don’t see any legal issues either – for the use of superprotect and for enforcing the Media Viewer, the opposite could be the case, superprotect should then be better used to stop using the Media Viewer for legal issues and complaints made off-wiki, so what happened here, is not logical at all. ;)

But fact is, that Deputy Directors are in the position to make office actions, that Erik is such a Deputy Director, that he described his edits as "WMF action" and as "WMF edit", so that is isn’t clear, if an office action was meant. And because of that, the edit in the Commons.js is now taken as an example for such an office action which I think is very confusing. And office actions are "only occasioned by a complaint made off-wiki (e.g. mail, email, telephone calls or personal meetings; but usually written) to the Wikimedia Foundation about the content of a Wikimedia project". So, either it has been an office action, then there have to be legal issues about the content in any way. Or is hasn’t, then why did he write something about WMF actions and WMF edits, and then it has been a misuse of the office action policy. But then it shouldn’t be taken as an example for an office action and also the office action policy shouldn’t be changed because of this, which is, what Sicherlich seems to want. --Winternacht (talk) 12:44, 16 August 2014 (UTC)

This has to be clarified, so that we all talk about the same thing. In addition, there are also comments in the German Umfrage that the WMF is legally responsible for the website, and therefore needs the superprotect. This also means in consequence, that the superprotect is or shall be only used in accordance to the office action policy, and that the superprotection of the Common.js page has anything to do with that policy. So, as you see, there is much confusion about this policy and superprotect. See for example Wikipedia:Umfragen/Superschutz#Kontra Vorschlag 4:

• „--Krd 12:32, 15. Aug. 2014 (CEST) Der Vorschlag ist Unsinn, es muss die Möglichkeit geben, _nötige_ office actions stabil umzusetzen. Vielmehr ist zu klären, wie mit missbräuchlichen office actions umzugehen ist.“.

„Deshalb kann nur der Seitenbetreiber juristisch der alleinige "Hausherr" sein, dass aber sollte jedem klar sein. […] Fakt bleibt, die WP muss im Zweifel Ihr Hausrecht auch gegen deutsche Interessen und die WDME durchsetzen müssen, weil nur Sie im Zweifel für deutsche Edits haftet. Das ist leider rechtlich zwingend erforderlich und auch am Ende auch gut so. --95.90.195.197 17:20, 15. Aug. 2014 (CEST)“

There are many comments which refer to the office actions and legal responsibility. If there has been no legal reason, then why should people talk about legal responsibility of the WMF all the time? And what has the superprotect right to do with legal responsibility? --Winternacht (talk) 13:06, 16 August 2014 (UTC)

We are actually (and have been) in agreement on clarifying roles and accounts. Stay tuned for comments from staff on next steps. -- LilaTretikov (talk) 16:51, 18 August 2014 (UTC)

Strategy to put in place best practices for product development

I am looking forward to hearing more about this: Lila laid out her strategy to put in place best practices for product development. We will communicate sooner, we will prioritize smarter, we will test more, and we will achieve better outcomes. Her vision is to involve the community at each step of product development, including more structured feedback stages and reviews from the recent Board meeting [18]. Deltahedron (talk) 17:25, 14 August 2014 (UTC)

This is what interests me as well. THIS is what i want to hear, this is what was already improving and needs a whole lot more improvement. The rest be drama. TheDJ (talk) 06:59, 15 August 2014 (UTC)

It's nice sounding words at the moment. If the first act after delivering this strategy is to "stick it to the community" by supporting the most crass power grab I have seen in the Mediawiki sphere, then the rest is not merely drama, it is tragedy. Rich Farmbrough 18:05 15 August 2014 (GMT).

This is under work currently as a priority. It will be available for everyone's comments and suggestions. -- LilaTretikov (talk) 13:17, 17 August 2014 (UTC)

Status: on the road

Well ... we all hope that you are not lost on the road  ;-) nowhere a sign[19] of life ? Have a nice weekend :-) Best --Gruß Tom (talk) 08:54, 16 August 2014 (UTC)

Let's try and be a bit more considerate here Gruß. People do have personal lives and family commitments. Lila is still new and probably has a lot on her schedule. She was at least here reading and replying. The board and the architect of this issue are still relatively silent, and have been since this started. Theo10011 (talk) 11:28, 16 August 2014 (UTC)

No problem. It might be a shock to be confrontated with a community which never sleeps ... but this is a strenght and a part of wikipediaculture. It should be nice to read a notice from her latest on tuesday. BTW further reading: the press[20] and the votes[21] in de:WP. Best regards --Gruß Tom (talk) 12:46, 16 August 2014 (UTC)

I am not lost but I'd like to be.. sometimes ;) Remember there is only one of me...-- LilaTretikov (talk) 00:36, 17 August 2014 (UTC)

Thanks for your response. We do understand that it is not easy to deal with the current affair. You extended your announcement of (semi-)absence for one week. That you (or somebody) changed the announcement of your absence which is no longer visible[22] in this page history is not fine. If your house where on fire you could try to return ... or read your insurancepapers. Well it is your choice. Regards --Gruß Tom (talk) 08:32, 17 August 2014 (UTC) P.S. to those who operated on historyversions of this page: “You may fool all the people some of the time, you can even fool some of the people all of the time, but you cannot fool all of the people all the time.” (A. Lincoln)

No, that’s normal. [23] It’s about privacy, and it doesn’t matter here, where she is. No one has to reveal his or her own IP address, and this surely happened by instance. So I don’t understand your comment on this, privacy has nothing to do with fooling anyone or houses on fire. --Winternacht (talk) 19:46, 17 August 2014 (UTC)

Lila’s questions:

All – I am still traveling and will be in low-bandwidth areas until Friday. I wanted to drop you a quick note -- so you know I am not ignoring this -- it is top of my mind. I am reading your comments. I care and am sincerely sorry for any hurt feelings stemming from this. We clearly opened up a really big issue that needs resolution. I don’t want us to rush. I want to understand this and think it through and to work together to make this work. We need to figure out how.

I put together a list of the questions I am struggling with, or confused by, that you can shed your perspectives on. We may disagree in the end :) But I hope we find a way. They are roughly grouped by random subtitles, but really there is no particular order here. Feel free to add your own.

Decision-making

• How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?

• How can we reach representative decisions for all wikis? Good practices in product tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?

• Do we think RfCs to be the best means of achieving a decision or consensus? -- They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?

• Does every decision need to be community vetted? If not, where is the threshold?

• Does every decision need to be WMF vetted? If not, where is the threshold?

Communications

• Where should we communicate to reach enough critical people so we don’t get into a place where we waste work? Are we using the right communication channels to reach contributors? Readers? How do we prevent ourselves from jumping in at the very last moment (if we missed getting engaged on-time) when the works is ready and public?

• Who should be a part of the communication distribution? Should it be exclusively the Community, the WMF, or a combination of the two? If a combination of the two, how should responsibilities be broken down and shared?

• Why are personal insults a part of our methods of communication? How do we stop that?

• How do we stop pointing fingers? How do we align ownership, execution and authority without blocking all changes?

Planning for the future

• Is it possible for us as a community as a whole (and as many smaller communities) to think beyond what we do today: editing long-form Wikipedia articles? If so, how would we do this? I'm not saying this because I don't think we should work on making long-form editing easier, I want us to recognise that there are more types of content and ways to contribute and consume it today.

• What external threats should we be planning for?

• Practices in Technology

• WMF is a technology org. Do you disagree with that mandate?

• Our current change-management practices around Site Operations/DevOps are dangerously out of date (especially how we change production code). How should we fix that?

• What did you like that WMF development has done? What parts of the process worked? Which parts of the process did not work specifically, and why?

Readers & Editors

• To what degree should we care about the readers? Reach?

• Should we split the site into reading and editing experience? New users and contemporary?

• Do we believe that the expectations of both readers and contributors have changed over the last 10 years? Our UX studies show this to be the case: the users who come to Wikipedia today get confused because they have been trained to expect different UX patterns. How should we handle this?

• Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?

Features

• Do you think old features need to be redone or removed? What would you do to clean up old cruft and features? How would you tell the few hundred editors that would get really upset?

• What are the top 3 product features that you would like to see rolled out over the next two years for readers, assuming we could meet your quality standards?

• What are the top 3 product features that you would like to see rolled out over the next two years for editors, assuming we could meet your quality standards?

-- LilaTretikov (talk) 00:36, 17 August 2014 (UTC)

WMF

As you requested here some additional questions:

• How can we avoid WMF developper staff to insult regularly local communities, why is there an attitude from the paid staff of beeing superior to volunteers work and their decisions? What are the reasons of unpatience and inappropriate communication skills by leading devs?

• Comment: As recently as Wikimania there was a discussion on a Culture of Kindness. I think that sometimes we all forget that, but it's something we can all work on. -rachel diCerbo (WMF) (talk) 16:00, 18 August 2014 (UTC)

• By which change of structure/communication is it possible, that issues, which are important and vital to certain communities, can be earlier, more respectful and without empty marketing phrases could be adressed by the WMF?

• Comment - We at the WMF have been discussing more effective communication and collaboration with communities and are working towards that. The reason Lila is asking these questions is because we need your help. -rachel diCerbo (WMF) (talk) 16:00, 18 August 2014 (UTC)

• How do we avoid, that longtime productive, engaged, project-loving and decent personalities (like de:Benutzer:Drahreg01 recently and many others these days) are regularly leaving the project because of WMF issues?
• How do we manage the aim of project-wide votings concerning new developments without banning user from full participation, which are not able to communicate properly in english or are to shy to do so? Would it be helpful to provide official strategies, plans, developments and statements by the WMF with official translation into the languages of the communities (perhaps the 10 biggest WPs)?
• Is it possible that the WMF board defines, under which terms and conditions (number of participants, right to vote etc) the vote of a community is in their opinion an express of their will (or not) as it is mentioned in the guiding principles stated by Jimmy Wales? (btw these principles are still an inspiration and one of the main reason for volunteers supporting this project over the years?
• How can we roll back the influence of financial and marketing aspects on decisions of the WMF, which are increasingly discouraging volunteers?

--Gleiberg (talk) 01:34, 17 August 2014 (UTC)

Comment: hazards of question choice

I've seen a series of questions like this (or, these) lead things astray, by imposing an artificial structure on the discussion that misses important overview about the situation.

For example, the questions about the importance of readers. The wikimedia community contains a large number of highly intelligent people who started out as readers, and therefore can reasonably be expected to have great insight into the process by which readers become contributors. In fact, I'd have to say the community should contain a large group of people collectively much more qualified than the Foundation to make sound judgements about how best to draw in new contributors. How to make decisions should therefore translate as, how to tap into this vast potential wisdom of the community as a valuable resource (rather than trying to put a process in place to placate the community by making it feel respected; that would be an appalling waste of an available resource). Tapping the potential of the community does take careful thought; you have to consider what a wikimedian is likely to be able to notice. I'm reminded of something (half-remembered) from The Hitchhiker's Guide to the Galaxy, where the Vogons, in demolishing the Earth to make way for an interstellar bypass, say the Earth's inhabitants should have known about it because there'd been a notice posted at Alpha Centauri for several years. --Pi zero (talk) 03:28, 17 August 2014 (UTC)

Agreed. The reader doesn't exist. See Stupidity of the reader. --Nemo 22:23, 18 August 2014 (UTC)

Valid points, and thank you for them Pi zero. I've passed on your comment to the Growth team. As for the Hitchhiker's reference, you have hit the nail on the head about where information is listed. Do you have any ideas for resolving this? -Rdicerb (WMF) (talk) 23:37, 18 August 2014 (UTC)

Here is a real world example for the hitchhiker scenario: Starting with the week of 25th of August the default size of thumbnails will be increased from 220px to 300px. In the German wikipedia the first note on this change was issued a few hours ago with a single line on the little read page de:Wikipedia_Diskussion:Projektneuheiten ("Upcoming changes"). This will change the look of articles. Authors who have optimized the amount of images in "their" articles might not fall spontaneously in love with the idea. The only way to know about this change more than seven days in advance, would have been to read a page at Alpha Centauri, err, http://wikitech.wikimedia.org/wiki/Deployments.

In every major language version of the wikipedia there are places where authors hang out and share ideas. In the English wikipedia this would be the Village Pump or the Signpost. In the German WP it would be the Kurier. Spot the place that is most popular and post notes on planned changes. Communicate the plans as a proposal, not as an announcement. Be sure to read the comments on the discussion page. Give authors time to adapt their mindset to the upcoming change before the change happens. And be prepared to change your plans if valid arguments are raised. Yes, it is that simple.---<(kmk)>- (talk) 03:20, 19 August 2014 (UTC)

I'd certainly agree that's an example of Alpha Centauri; but while the fora you mention would presumably be less bad, I don't entirely agree it's that simple.

At en.wp, anyway, I've tended to find the village pump more of a firehose, and whenever I've added it to my watch list I've removed it because I simply couldn't keep up with its output. That's less Alpha Centauri and more like How to Hide a Hippopotamus. (There doesn't seem to be anything about that old children's book on en.wp; the answer, on the last page of the book, is that you hide a hippopotamus in a herd of hippopotamus(es/i).)

As for Signpost... at en.wn we sometimes refer to it as Slimepost, as we see it as running consistently biased coverage of Wikinews with the de facto editorial stance Wikinews delenda est. I for one didn't even try to comment on the recent "op-ed" there pretending to be an exposé of en.wn (I'd rate the piece as creative fiction), because I thought it a waste of time and likely source of pointless pain to attempt rational discourse in the middle of a rabid Wikinews hate-fest. The upshot is that each time a Signpost comes out, I usually only do a string search for "wikinews", so at least I'll be aware when they decide to slime us again, and otherwise I don't read it. (That does bring up the problem of toxic social atmosphere on Wikipedia... a problem I believe is a long-term effect of subtle factors in its social infrastructure... but that's an incredibly difficult systemic problem, and not what we're discussing here. Though I perceive all these different problems to be interconnected, in one vast tangle.)

So both of those fora have their problems, too. The firehose/hippopotamus problem takes me back to Rdicerb (WMF)'s question, about ideas for better communication. My thoughts are still pretty vague, but I have a feeling this is somehow a technical problem (if you've got a hammer, everything looks like a nail) — somehow it ought to be possible to help people match themselves up with specific fora where things they're interested in... or should know about, which greatly complicates things... are being discussed. I'm apt to spend years waiting for inspiration on these sorts of things, because my standards are pretty high; I'd want something blindingly easy to use and highly effective that wouldn't in any way increase the technical burden on users and would, in fact, just seem to be a natural part of the scenery. Perhaps some sort of very short automatically generated summary list of stuff of interest? Site notice type stuff doesn't seem quite right, user-talk notes would accumulate, email is even more cumbersome. And what to do about people naturally tending to opt out (or tune out) of hearing about things and then being upset when they don't? It seems the whole problem has been concentrated but not solved. --Pi zero (talk) 11:16, 19 August 2014 (UTC)

If you post on WP:VP(T) I too will probably miss it because I am busy with other things. But a lot of other people who are "basically sound" will see it, and I am (reasonably) happy to put my trust in them, since I can't watch everything. Rich Farmbrough 13:03 19 August 2014 (GMT).

Comment to one of Lila's questions

decisions that move us forward, rather than backwards - this question leads to another one that is much more essential: where is forward and where is backwards? where should wikipedia move to, and why? It is this question that needs to be discussed but at the time being is ignored. So it is hardly possible to say whether we agree or disagree and why. This is one reason why the media viewer has been treated as a pure question of power (especially by WMF). In the media viewer case, it should simply be acknowledged that at the core there is a dissens between "design" (in the sense of aesthetic presentation) vs. "information" (for use and understanding of media files). I, for my part, feel that information in this sense is a vital part of an encyclopedic project whereas aesthetic presentation is nice to have. The WMF developers seem to think that it is the other way round: information is nice but negligible whereas aesthetic presentation is vital.--Mautpreller (talk) 10:30, 17 August 2014 (UTC)

In addition: one might say that these two objectives need not exclude each other. I utterly agree. However, a software change that (disputably) might improve aesthetic presentation at the expense of vital information for reuse and understanding of the media files does treat these two objectives as mutually excluding each other. I feel that this is fundamentally wrong. And it is even more wrong to "solve" this problem by force instead of bargaining.--Mautpreller (talk) 11:20, 17 August 2014 (UTC)

Comment: Hi Mautpreller, I hope it's OK that I jump in here. Something that's important to move towards is better collaborative communication, more structure and clarity. Without getting into the specifics of any particular product, moving towards more effective partnership between communities and the WMF is vital to moving forward. You ask where Wikipedia should move to, and why? In my humblest opinions as a new staffer at the WMF: What this community has achieved is incredible. The world's repository of knowledge, created in partnership with developers (both staff and volunteers) and content creators worldwide, is an example of true collaborative partnership. It is not without conflict, but I hope that our common ground is that we all care about this very much, which is why there is this tension right now. I'm definitely oversimplifying things but hope that by starting from there we can start to define where "forward" is. -Rdicerb (WMF) (talk) 19:41, 18 August 2014 (UTC)

Dear Rdicerb, that's a friendly answer but it does not really help me. Top-down enforcement of an inappropriately designed tool by means of technology (such as WMF has done in this case) is definitely not in the frame of "true collaborative partnership," it is exactly the opposite. How do you think that trust can develop between the WMF and local communities after such a measure? Enforcement will spawn suspicion rather than trust.--Mautpreller (talk) 12:47, 19 August 2014 (UTC)

Answers by Müdigkeit (talk)

Thank you Müdigkeit for your thoughts -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Decision-making

How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?

Answer: That is difficult- but please note that in this situation, lots of authors asked for a delay until certain issues are fixed.

Comment: This kind of feedback needs to happen, but earlier in the process. Multiple issues occurred with MV that we need to learn from. However when the software is post-production, it needs to be iterated and those issues need to be fixed in-place. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

How can we reach representative decisions for all wikis? Good practices in product tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?

Answer: The time is important. My suggestion is either: alpha->first beta with opt in-> bugs fixed and some time->opt out for registered users to let them test that-> in case of critical flaws, revert, else fix bugs and release->fix important bugs afterwards. If that isn't possible, and you have to implement it on all, revert features that have flaws like breaking pages, breaking the license...

Comment: We are planning at something quite similar. I think you may know that there was an opt-in Beta phase, but one of the admins removed the Beta link. This is an issue with the in-prod changes. We are also looking at incremental roll-outs 5%-10%, etc. Reverts need to be consistent -- they are an either all or nothing game. Otherwise it becomes and expensive management nightmare. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Do we think RfCs to be the best means of achieving a decision or consensus? -- They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?

Does every decision need to be community vetted? If not, where is the threshold?

Answer:Decisions that do not effect the writers community... just do it. If it effects the writers only- be careful. You don't want to lose them over some broken or simply not wanted feature. Some ideas sound good but aren't. Decisions effecting the writers and the readers are dangerous if they help one side and harm the other. Bad readability for a writing improvement is obviously a bad idea. But decisions harming the writers are dangerous, too. No editing community->wiki gets worse-> bad for everyone. Communicate closely. And listen to feedback. Be carful- and do not rush anything to release.

Comment: This makes sense. The problem is, like with MV a reader feature can be a sticking point for the writers. We have very opinionated users. I think we need to attempt to not degrade the writer experience at the minimum with reader features. It would help if the writers, however, gave some room for reader-focused experimentation. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Does every decision need to be WMF vetted? If not, where is the threshold?

Answer: The WMF cannot be everywhere. The WMF should focus on legal problems and software issues at the one side, but also stop communities violating the principles- that is, advertisements or something like that instead of neutral wikis, toxic environments because of personal attacks... but it should not edit every single article! The WMF should only do something if the community cannot fix it.

Comment: I think WMF needs to be the "foundation" for the community. This means raise funds, build software, deploy programs, do customer support (legal, etc.) that communities cannot handle. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Communications

Where should we communicate to reach enough critical people so we don’t get into a place where we waste work? Are we using the right communication channels to reach contributors? Readers? How do we prevent ourselves from jumping in at the very last moment (if we missed getting engaged on-time) when the works is ready and public?

Answer: The first step is having an idea. The second step is asking if the idea is good. Ask the community for feedback- if the idea has flaws, if it should be implemented at all. This step filters out bad ideas- that are bad ideas in general. And it can help to avoid possible problems. The third step is writing and testing code. Important changes to the feature should be handled like a new idea- first ask, then implement. When the product is beta- create/ use a small testing environment to look for bugs/problems. Before you deploy- inform the communities(you are doing that) and set the deploy date to a date where enough staff is there to fix bugs(important).

Comment: I think our major issue with communication is location. I, for one, am really struggling to find the right places to have even this discussion. We need a central hub and methods to reach people (like email or alerts). -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Who should be a part of the communication distribution? Should it be exclusively the Community, the WMF, or a combination of the two? If a combination of the two, how should responsibilities be broken down and shared?

Answer:I am not sure of the meaning of the question, but some general page where suggested changes and developed changes could be handled would be great...

Comment: I was wondering if community members can help reach others. Our "network" is vast, yet we are not leveraging its capabilities except for crisis. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Why are personal insults a part of our methods of communication? How do we stop that?

Answer: Block those who insult others personally for a period of time, if they don't stop, infinitely.

Comment: I am wondering the same thing, but today this something the WMF is careful about. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

How do we stop pointing fingers? How do we align ownership, execution and authority without blocking all changes?

Answer:Some will always point fingers. Always remind that Wikipedia is not usual business. Everyone understands that legal problems are your problems... and ours, if you ignore those and get sanctioned. Blocked changes- they could affect us in the harms the authors, nice for readers problems. Accessibility for readers is important. But for contributors, too... be extremely careful. You notice nearly nothing until the community drops below a certain point. Then... it is almost too late and the wiki begins to collapse, affecting everyone. Do not underestimate that. Again, not having the team ready to bugfix at release is an easily avoidable problem.

Comment: I am wondering if we can figure out how to problem solve together vs. assign blame... -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Planning for the future

Is it possible for us as a community as a whole (and as many smaller communities) to think beyond what we do today: editing long-form Wikipedia articles? If so, how would we do this? I'm not saying this because I don't think we should work on making long-form editing easier, I want us to recognize that there are more types of content and ways to contribute and consume it today.

Answer:No idea.

Comment: I think it would be incremental local innovation. But I think the WMF needs to be there to recognize it and support it. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

What external threats should we be planning for?

Answer: Forks, those who use our content and have an advantage(google, for example)

Comment: Yes on content use by known brands. Forks are much less dangerous, history shows that successful forks are hard to do. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Practices in Technology

WMF is a technology org. Do you disagree with that mandate?

Answer:Maybe no.

Comment: I don't think the community as a whole is set up to build software. The community is built around building content. So software is utility. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

This is not correct. The software (MediaWiki) was built by and around the community (which was made around, for and through the content and the wiki-ness). When I think of my personal "top 3", I can't find anything done or doable by the WMF; usually volunteer devs make most positive impact on my use of the wikis. --Nemo 22:23, 18 August 2014 (UTC)

Our current change-management practices around Site Operations/DevOps are dangerously out of date (especially how we change production code). How should we fix that?

Answer:I do not have sufficient knowledge in programming

Comment: I come from a background of operating large web projects. In any internet site of our size (even community-based ones) site files would have been only accessible to system operations -- vetted individuals with background checks. What I've missed here is that a standard practice elsewhere -- here was thought of as a violation of editor rights. That was surprising and not intuitive to me. What I am understanding now (and this is still a forming) is that a change like this needed to be communicated through the community first. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

What did you like that WMF development has done? What parts of the process worked? Which parts of the process did not work specifically, and why?

Answer: Beta worked- on the side of WMF. The early questions about the features didn't happen, but that might have been not a major problem. The release itself, however: too early deployed for unregistered users(and that means: nearly all readers), not enough staff ready to fix bugs after release and to react to comments after release(I might be wrong here, but deploying features before sunday is never a good idea, better wait a bit)

Comment: I think we have a lot of good components, but not a full recipe yet. We can disagree on the readiness of MV, but I think if we quantify "ready" as X % opt-in rate it would be much harder to argue. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Readers & Editors

To what degree should we care about the readers? Reach?

Answer: Not losing readers and getting some is certainly important. Accessibility is important.

Comment: I would go further and say we need to be gaining. Especially in growing regions. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Is our mission to move from top 5 website to top 4 or top 3? No, it is to make knowledge available. Simple pageviews are not a sophisticated measure, see my response to this question below. Rich Farmbrough 13:08 19 August 2014 (GMT).

Should we split the site into reading and editing experience? New users and contemporary?

Answer:ehh... sounds like a bad idea. But I am not sure.

Comment: This is an extreme solution if majority of the writers just are not willing to accept change -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Do we believe that the expectations of both readers and contributors have changed over the last 10 years? Our UX studies show this to be the case: the users who come to Wikipedia today get confused because they have been trained to expect different UX patterns. How should we handle this?

Answer:Expectations change, and we have to change. But in the right direction- still preserving our goals.

Comment: Agreed -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?

Answer:No,No. Our mission is to provide content for readers. But without sufficient editors this mission fails. And we do not have that many...

Comment: Thank you. I think editing is changing too -- people are getting new forms of interfaces and editing is becoming much more iterative, short form. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

Features

Do you think old features need to be redone or removed? What would you do to clean up old cruft and features? How would you tell the few hundred editors that would get really upset?

Answer: If they only hinder development and such... remove that. But don't do it in rush- tell them that it will get unavaliable soon, and why you cannot keep their feature.

Comment: We have some that are very very old, including inconsistent old templates, but some people just love them. It's expensive to maintain. I like your communication idea. It makes sense. I still think people will get upset. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

What are the top 3 product features that you would like to see rolled out over the next two years for readers, assuming we could meet your quality standards?

Answer:1.A better search engine(There is something like this in beta already, isn't it?), 2.+3. no idea.

Comment: Thank you. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

What are the top 3 product features that you would like to see rolled out over the next two years for editors, assuming we could meet your quality standard

Answer: Actually... nothing. But there might still be features helping other editors, maybe simplifying some processes for new editors...

Comment: Today we are turning away new editors at a rate of 99+:1. AS an example 9 out of 10 people with immediately leave the edit page as soon as they see it. This is scary. This is before they even attempt to edit, get reverted, etc. We have to fix that new editor onramp, IMO. -- LilaTretikov (talk) 15:00, 18 August 2014 (UTC)

--Müdigkeit (talk) 16:59, 17 August 2014 (UTC)

Some answers from a "reader" who has followed Wikipedia for quite some time

Thanks for your note -- you should create an account so we can invite you to test -- LilaTretikov (talk) 15:09, 18 August 2014 (UTC)

How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?

Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?

"Forwards" and "backwards" are a matter of opinion. In the context of Media Viewer, I'm reading this to mean that "backwards" means reverting the software and "forwards" is keeping it in use. But, this is a matter of perspective. I see Media Viewer as a massive step backwards in performance, functionality, and aesthetics. The English Wikipedia, German Wikipedia, and Wikipedia Commons communities seem to concur with my opinion. Only the community that is the WMF, which has a significant conflict of interest given it probably spent hundreds of thousands of dollars developing Media Viewer and burned an incredible amount of goodwill, disagrees. Don't fall into the sunk cost fallacy!

But to get beyond your loaded question, the way we make decisions should be rather straightforward. You seem to misunderstand the role of the WMF, which is particularly dangerous as you're the Executive Director of the organization. You apparently see the WMF's mission to be the mission of the Wikimedia movement as a whole. That's wrong. For a projects like Wikipedias, it is useful to have a corporate entity to raise funds, spend funds, run technical infrastructure, deal with legal issues, etc. Given that the community as a whole cannot perform these functions, the WMF was created to serve the community in these ways. The community is the organization that actually is responsible for creating and improving vast bodies of knowledge and spreading them around the world. Which leads to your next questions...

Comment: What I mean by forward is improving on the baseline, vs. abandoning. I also mean keeping up with the user experience changes that are happening in the digital world around us. Some things are worth abandoning -- sometimes we will make something that does not work out to be as useful as we thought. Throwing code away is part of the process. But in order for us to find out we need time to watch it live and measure. Today's RfC process is not effective for that. -- LilaTretikov (talk) 15:09, 18 August 2014 (UTC)

Does every decision need to be community vetted? If not, where is the threshold?

Does every decision need to be WMF vetted? If not, where is the threshold?

Do we think RfCs to be the best means of achieving a decision or consensus? -- They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?

How can we reach representative decisions for all wikis? Good practices in product tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?

Ultimately, the decision-making authority for everything comes from the community. A good deal of operational authority has been delegated to the WMF because to do otherwise would be unproductive micromanagement, but if the community takes a position, you have no business opposing it. The WMF is a servant, not a ruler. The only exceptions arise when legal considerations require changes that might go against the wishes of the communities, else the project itself may be compromised by outside forces.

Quite frankly, it's entirely irrelevant what the WMF thinks of how the community makes decisions through consensus, be it RfCs or otherwise. The community is entitled to make decisions in whatever manner it sees fit as long as there is strong consensus about those methods. It's certainly not wrong for the WMF to suggest better methods of reaching decisions, but it's ultimately up to the community to adopt them. That's the wiki way.

It looks like we're in a position where the WMF has decided they know what's good for the mission better than the community responsible for carrying it out, and thus they will take technical measures to override the clear consensus of the community, despite their conflict of interest. The WMF has exceeded their delegated authority. The WMF has, at least where this particular issue is concerned, ceased to serve the community. This is completely unacceptable and it must stop. --98.207.91.246 17:14, 17 August 2014 (UTC)

Comment: I disagree with that. We don't have one community, we have hundreds. As the entity that supports ALL communities we need to have a baseline that is acceptable to all. So the process is important. And in the areas where the WMF has competency and expertise, like legal or fundraising, and increasingly software, it needs to lead those conversations. WMF is a part of the community as a whole that has a set of very specific jobs to do. It does not mean ignoring the other communit(ies), opinions and ideas, however. -- LilaTretikov (talk) 15:09, 18 August 2014 (UTC)

Features

Thank you Salix alba -- LilaTretikov (talk) 15:12, 18 August 2014 (UTC)

I only have one feature I really want to see: a better maths renderer. See en:User talk:Jimbo Wales/Archive 162#WMF plans for mathematics. The code is pretty much there but keeps stalling at code review. A small bit of hand holding from staff developers could be all that needed to get this through its blockage.

Comment: I know about this and would love to see it too, but I think it needs to go through the prioritization pipeline. -- LilaTretikov (talk) 15:12, 18 August 2014 (UTC)

For stuff to remove, nothing really stands out but there is a lot of small cruft about. For example there are several hundred thousand templates on en.wiki with much overlap and duplication. I'm sure much of this could go. Maybe there could be tools to help the process. In a similar vain is en:Category:Wikipedia backlog with some scary high numbers of little tasks to do. The same can be said policy wise, ten years of development has led to a large accumulation of policy. This can make it much harder for new editors to enter as they need to get up to speed on a large bureaucracy, probably the largest hindrance to attracting new people. In general I don't think its time for more features, more a time for simplification.--Salix alba (talk) 17:45, 17 August 2014 (UTC)

Comment: Unfortunately there is A LOT of cruft there. And removing some of it will cause some people to get upset :( but we need to do it as we update the platform. -- LilaTretikov (talk) 15:12, 18 August 2014 (UTC)

Some answers

Thanks Martijn Hoekstra -- some of my comments inline. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)

I replied again to some. I indented your comments, feel free to revert that if you don't like it.

A question you didn't ask by the way, but I still would like to answer, is how would you like to see the role of the WMF. I think the WMF could be invaluable as a facilitator. An organisation that helps the rest of the community realize their goals. Don't do software developement, but help the community do software development. Reach out with training. Practical ideas: develop some online classroom-style course, and divert resources from development to that. Provide some amount of hand-holding, for example organise bug squatting days not to squat bugs, but to take someone who is new to MediaWiki developement, and squash a bug together with them. If a community has technical needs (for example lua-template related) organize for help that can help them do it (which is something else than doing it for them). Tool labs is a project that takes steps in that direction, which is great. Martijn Hoekstra (talk) 18:21, 18 August 2014 (UTC)

Decision-making

• How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?
  • How do we determine which change is a change forward? I think that's the central question, as nobody(tm) actively wants us to go backwards.
    • Comment: Forward is two fold: catching up with current tech, and building on top of current. Most communities naturally gravitate towards it's own point of gravity and become resistant to change. This is normal. But it is also very dangerous as it is insulating. So this is a question about instigating change. And any change (net positive or net negative) breaks some things in the process. In the end what really matters is innovation into the next decade. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • It's one of the reasons why I really regret not having any real competition. If we would have one, and someone does what we do - making information freely available to anyone - better than we, we still win. Everybody wins then. Up until that point we have to do it ourselves, and overcome some of our change-aversion. But the question as it's phrased is so broad there is no real answer to it. Martijn Hoekstra (talk) 18:11, 18 August 2014 (UTC)

• How can we reach representative decisions for all wikis? Good practices in product tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?
  • The individual wikis are diverse and pluriform. Not having every install identical is not a problem.
    • Comment: From perspective of both: user experience consistency and software maintenance costs it is a huge problem. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • I think user experience consistency is somewhat of a red herring. Generally, it's a good thing, but user experience consistency is also culturally defined. When we have no or almost no reader overlap between zh.wiki and de.wiki, it doesn't really matter all that much if the user experience is consistent between the two. The sister projects have different goals and very different presentations make sense for them. When it comes to software maintenance costs; yes, that's a problem. Empowering the local communities to take a large part of the burden of the maintenance upon them would be a step towards mitigating that problem. Martijn Hoekstra (talk) 18:11, 18 August 2014 (UTC)

• Do we think RfCs to be the best means of achieving a decision or consensus? -- They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?
  • It would be nice if we could come up with something better.
    • Comment: I agree. We need a way to have clear goals and understanding of what is "deployment ready" means and be quantitative. We are working on defining this and will open that up for comments. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • On an RfC? ;) I don't think the question on how to hold a good (online) comunity discussion and find consensus has been solved for the size of our communities yet. RfCs are a flawed attempt. I'm not sure if any alternative isn't at least as flawed. That doesn't mean putting some effort toward finding out is a bad idea. Martijn Hoekstra (talk) 18:11, 18 August 2014 (UTC)

• Does every decision need to be community vetted? If not, where is the threshold?
  • What does "vetted" mean exactly? Most small changes most community members don't mind, which means that - even if there is a majority against a change, it's ok to go through with it, if the opposition isn't too strong; a handful(maybe 10 or 20) of people at an RfC or something. But when a large part of the community starts screaming bloody murder, there is a large group of people participating at an RfC, there is a problem.
    • Comment: The real issue here is when. We must figure out how to get people on-board earlier. When the software is already on 790+ wikis, a large screaming is a huge problem, but is not very helpful. We need these same people to engage during the conception phase... -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)

• Does every decision need to be WMF vetted? If not, where is the threshold?
  • Again, this pushes in a direction where the WMF is not part of the community. That's terrible. Legal should definitely be WMF vetted though.
    • Comment: I think BOTH as ineffective: everything vetted by both. I think we need to agree where the line is and who bares the responsibility for what. That would give us efficiencies and freedoms to move faster. This does not mean we don't take each other into account, it's just means we have clarity around what we do. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • If the WMF is part of the community, community vetted means WMF vetted as well. "Everything that won't cause the projects to explode in a ball of fuming hate, pitchforks, and comically stylized speech bubbles filled with skulls, crossbones and bombs" should be a reasonable start for things that the WMF can do unilaterally without getting buy-in from the rest of the community. Martijn Hoekstra (talk) 18:11, 18 August 2014 (UTC)

Communications

• Where should we communicate to reach enough critical people so we don’t get into a place where we waste work? Are we using the right communication channels to reach contributors? Readers? How do we prevent ourselves from jumping in at the very last moment (if we missed getting engaged on-time) when the works is ready and public?
  • Anyone who said they didn't saw it coming is either lying or doesn't understand a thing about the movement. It also depends on the question asked. If the question asked is "how can we best tweak x" you forgot asking "should we do x", and if yes "how should we do x".
    • Comment: We need more engagement in the prioritization phase, yes. But I also agree that it's on all of us to communicate, not just the WMF. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • I think it's the burden of the person with the initiative to initiate the dialog. In case of WMF software projects, that's the WMF.

• Who should be a part of the communication distribution? Should it be exclusively the Community, the WMF, or a combination of the two? If a combination of the two, how should responsibilities be broken down and shared?
  • We're still calling the WMF a separate entity from the community. The more often you do that, the more true it becomes.
    • Comment: How would you phrase it and still be clear? -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • There is a reason the WMF is a disconnected entity from the Wikipedia community and that is a legal one, for the WMF as a sole ISP function for distributing (like websites) it has legal protection, if we want to keep it that way we have to stop the Cathedral approach where decisions are made for better in the cathedral and layed upon Wikipedia, the problem inhere, the American legal system might conclude that the influence of the WMF in shaping the content by modeling, introducing (or restricting) the available tools makes it more vulnerable to a legal attack on this issue, introducing a function to override the community increases liability i think.Mion (talk) 21:56, 18 August 2014 (UTC)
      • I generally phrase it as "the WMF and the rest of the community". Martijn Hoekstra (talk) 09:24, 19 August 2014 (UTC)
• Why are personal insults a part of our methods of communication? How do we stop that?
  • Because people are getting desperate. Prevent that.
    • Comment: I think no matter how desperate, people have to own how they behave. Insults are unacceptable and frankly damaging to the outcomes. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • Yes, all true. But let's start addressing that point when we're not in a mini-crisis like this.

• How do we stop pointing fingers? How do we align ownership, execution and authority without blocking all changes?
  • I don't think there is something specifically wrong with pointing fingers, as long as it's done constructively. How do we enable people to express their concerns, even if it is about people, respectfully and effectively is a better question. I'm failing to fully understand your second question
    • Comment: Your second question is what I am really after. I think pointing fingers is different from constructive critique. The first one is passing judgement and does not encourage people to get better, because it is final. The second engages people and makes suggestions. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)

Planning for the future

• Is it possible for us as a community as a whole (and as many smaller communities) to think beyond what we do today: editing long-form Wikipedia articles? If so, how would we do this? I'm not saying this because I don't think we should work on making long-form editing easier, I want us to recognise that there are more types of content and ways to contribute and consume it today.
  • Good to see the we are a community again. Possibly. This is really the first thing I'm hearing about it.
    • Comment: I think it is a hard one to solve without leadership coming from somewhere... -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)

• What external threats should we be planning for?
  • Cencorship and advocacy.

• Practices in Technology

• WMF is a technology org. Do you disagree with that mandate?
  • Somewhat.
    • Comment: Can you elaborate? -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • Yes. To me, the WMF is the organisation that runs the - well - organisatorial side of things. Legal, payment processing, and the organisation that supports its own organisation (HRM, ED's). That a large part of the software development is done by the WMF happened almost accidentally. People worked on MediaWiki. The foundation had the money; it was only natural and fair to employ those people, to let them work on the product full time. This effectively - and without any malice or bad intent - created a divide between the developers who are WMF on the one side, and the rest of the community on the other. With the communities traditional dislike for payed content work, a programmatic and content side of the WMF never developed. Thus a technology focused WMF came to be without that being its original mandate. That became the center of gravity around which the WMF mainly evolved, and borrowing your own analogy, attracted more weight. Technical tasks that are now completely the WMFs - who is responsible for hosting (which is a question separate from who pays the bills), networking, ops, who has the keys to the servers - could have been the general communities rather than the WMFs. That's not how it happened, and this is where we are now, where the WMF is a technology org, but I think that has more to do with historical accident than with mandate. Martijn Hoekstra (talk) 18:11, 18 August 2014 (UTC)

• Our current change-management practices around Site Operations/DevOps are dangerously out of date (especially how we change production code). How should we fix that?

• What did you like that WMF development has done? What parts of the process worked? Which parts of the process did not work specifically, and why?

Readers & Editors

• To what degree should we care about the readers? Reach?
  • I'm part of a small minority that says screw the readers. That's an odd viewpoint since our mission has always been to give our content to the public. That has also always been our weak suit. Our strength has always been content creation. Wikipedia worked in the first place because it had content, rather than presentation. Due to our licence we're the perfect vessel for re-use. Any compromise where content creation and curation is hindered, even if it gives our readers something great in return is IMO a bad idea. Focussing on re-use, or even setting up separate readers projects would be more effective in the long term in my opinion. I'd like to stress again this really is a minority viewpoint.
    • Comment: Our challenge is that the "Readership" mode is changing. If before people were coming to the site, they are no longer doing so as much -- the content is embedded. This means we cannot convert them to editors. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • Not as easily, no. Cooperation with re-users and having a good relationship with upstream (us) can mitigate some of that though, and it's in the best interest of longer term re-users as well. If they understand that, and help sending content consumers upstream, that can be a plausible and productive cooperation.

• Should we split the site into reading and editing experience? New users and contemporary?
  • well, see above. Possibly, yes. But we should take care that that doesn't compromise readers becoming editors.

• Do we believe that the expectations of both readers and contributors have changed over the last 10 years? Our UX studies show this to be the case: the users who come to Wikipedia today get confused because they have been trained to expect different UX patterns. How should we handle this?
  • Duh. We should deal with it on a case by case basis. Having intuitive UX is important. Sticking with our values (open access, anyone can edit, copyright is important and so is attribution) is more important. Better UX while retaining the latter would be fantastic. Worlds greatest UX while compromising on the latter isn't.
    • Comment: I don't think our values and contemporary UX are mutually exclusive. -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • And mostly they are not. But if we take a look at where it goes wrong - currently MediaViewer - it goes wrong because people get very upset over its compromises to those values. Not because they "don't like it" I think a great majority of editors likes the MediaViewer - just not the parts that endanger clear and open licensing information and attribution. If it goes (badly) wrong, it's generally because those values are being violeted. It hits us in the feels. Martijn Hoekstra (talk) 18:11, 18 August 2014 (UTC)

• Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?
  • No, but it primarily is. The quality of the content will always be more important than the quality of the UX. Buzzfeed has great UX.

Features

• Do you think old features need to be redone or removed? What would you do to clean up old cruft and features? How would you tell the few hundred editors that would get really upset?
  • backwards compatibility is a large problem: it's hard, sticky, and un-fun. Nobody wants to think about it and work on it, but it's imperative to our mission and tradition: the tradition of conserving knowledge. Wikitext for example should really go the way of the dinosaur (because it's horrible), but it is what we have now, and it's really mature. Replacing it with something new that could be great in the future will break a lot of stuff. Clear communication cycles are important, and communication is a two way process. If we were to say "we think wikitext is horrible, and we're thinking on replacing it. What conditions should be met before we can start thinking about that" would be great. "here is a prototype that will eventually replace wikitext. How can we tweak it?" not so much. Wikitext is the most extreme example, but it goes for everything.

• What are the top 3 product features that you would like to see rolled out over the next two years for readers, assuming we could meet your quality standards?
  • (loading all pages by default into) a visual editor: anyone can edit. I would like to see more video media, but I think the problem mainly is our lack of free video media rather than that they're not ideally presented. Possibly better interwiki.
    • Comment: What do you think happened with the video RfC? -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC)
      • I'm not sure honestly. Martijn Hoekstra (talk) 18:11, 18 August 2014 (UTC)

• What are the top 3 product features that you would like to see rolled out over the next two years for editors, assuming we could meet your quality standards?
  • (loading all pages by default into) a visual editor: anyone can edit. Non-linear edit-histories to facilitate splitting and merging articles. Improvements in the diff and edit conflict engine (maybe a split between article features, where changes are on the word level, and paragraphs are sometimes moved, and talk, where edits generally are new section insertions).

Comment: Thank you! -- LilaTretikov (talk) 15:47, 18 August 2014 (UTC) Martijn Hoekstra (talk) 18:48, 17 August 2014 (UTC)

• and you, for taking the time and effort to engage in this despite your travel difficulties. These issues are high-importance, and I appreaciate you taking the time right now to address them, even if they're (clearly) not being fixed tomorrow. Martijn Hoekstra (talk) 18:11, 18 August 2014 (UTC)

Answers by Sicherlich

Thanks for these Sicherlich - I have some thoughts inline. -Rdicerb (WMF) (talk) 17:23, 18 August 2014 (UTC)

thanks Rdicerb! I answered them ...Sicherlich ^Post 02:03, 19 August 2014 (UTC)

Decision-making

• How do we make decisions that move us forward....

Define where you want to go. Check if the communities agree in general with the way. (Please; not 99 goals. Nobody will read them and more important you cannot follow them anyways; 3 maybe 4. Thats more than enough.) Than show that you really go forward not sidewards, backwards or stepping on the same level. ... Measure it or forget it

Comment - Broad goals may even be disagreed upon across a wide scale, but I see where you are coming from. I think asking if the communities agree in general is something that will be impossible to gauge, as the communities are all different and have different cultures. As for "Measure it or forget it" - I think that the WMF and the communities alike tend to like that idea, and it's our hope that we will be able to utilize data in more effective ways in the future. -Rdicerb (WMF) (talk) 17:23, 18 August 2014 (UTC)

I worked for a large international NGO - we had common goals. Of course not everybody liked them to 100 %. That is IMO not necessary. Every board has its main goals. As long as they in general go with values and the vision people will accept them. Of course you have to communicate why these are your main goals for the year (or maybe for 5 years) and for people interested; what ar other point you see but, sadly, you have to put down for later ...Sicherlich ^Post 02:03, 19 August 2014 (UTC)

• How can we reach representative decisions for all wikis?

simple: you can not. There are so many cultures and you want to put them all under one rule? Wikimedia promoted diversity for the last years (okay, diversity was the nicer name for women aid, but anyways) but if it really comes to diversity you don't want it? Afraid of it?

• Does every decision need to be community vetted?

yes and no. if it is a bug; change it :o) - if there is some change; communicate it. The larger the change, the more communication is needed. … and stop defining right by "WMF thinks". Thats what you do at the moment with the super hero tool

• Does every decision need to be WMF vetted?

no. as most what WMF does has low effect for the communities anyway.

Communications

• Where should we communicate to reach enough critical people

depending on the projects. Therefore I thought you have the community advocates`? (Okay, the one on de seems not active there at all. So I might be mistaken). And communicate in the language of the people. The WMF has a problem with the german community: the discussion is mostly in english. One may assume that there is lot getting lost .oO Among other things it puts WMF in the stronger position what leads for sure not to an improvement of standing of the WMF on de (for other languages the same issue).

Further more we discuss here on your private talk page. Not that we don't discuss it on de: no its not even on a regular meta-page

Comment - There are Community Advocates in the Legal Department and for product development specifically there is the Community Engagement Team. As for me personally, I am also new to the Foundation and these communities, though I am no stranger to worldwide volunteering communities. As far as staffers who are connected to communities who do outreach, this is fairly ad-hoc and not always the most effective way of reaching the people that we want to (which, especially when a product or feature is going to launch, is the entire community). Please know that we want to work on this but there are blockers, such as SUL finalization. What do you think would be effective means of reaching people? You and anyone else reading are welcome to reach out to my team (Community Engagement) at any time with suggestions. -Rdicerb (WMF) (talk) 17:23, 18 August 2014 (UTC)

But I (lets asume me as the community even I'm dont intend to be representative for the community! :) ) dont have the urge to communicate :D - as long as you leave me in my undisturbed :o) - it changes in the moment you "disturb" me by making changes. People don't like changes and they hate sudden changes. So tell them before (see below I wrote about it: big changes; lots of communication; small ones: less communication :) ) ....

for your page some suggestion: translation. at least some main languages?! As well on your pages: which languages do you speak? Only Whatamidoing has a bubble box (stating more or less english only :/ )

"welcome to reach out to my team (Community Engagement) at any time " - how do I know who and where you are? :o) - first time I heard about you. ... And I'm not at all new to WP. What about people less then 10 years in WP? what about people who do not speak english and have no idea about meta? ... ::::I'm sure there are a lot of things to do but "I" (as Wikipedian) don't know about it and finally I dont care. You disturb me (with a change :D ) so I'm upset. Thats the issue :) .. maybe a different aproach helps as mentioned (above? below? ;) ) - sell the idea first.

"I am no stranger to worldwide volunteering communities" - that is very good. The difference here might be; some of the key people do not speak english. There are people who play an important rule on some wikis who do not feel comfortable with english. We loose theire input

"What do you think would be effective means of reaching people" - lowering the wikis I have to check. Super man issue on de: talking on de, then meta and now beeing told there is a team I can find on mediawiki.org - WOW! - adding the different pages on the wiki its getting realy wired :)

...Sicherlich ^Post 02:03, 19 August 2014 (UTC)

@User:Rdicerb (WMF): came to my mind earlier today; I wonder why in this communication issue Wikimedia Deutschland is not involved?

I don't know about the structure of Wikimedia too much I have the impression a lot of people do a lot of things which do not have any effect on any Wiki . But if WMF is really interested in communication with the communities why not put the chapters in place? They do speak the language, they know the culture, they life in the same time zone, they can easily go for some real life meetings. I remember we had this kind of thing in WMD but as far as I know they stopped doing it some years ago. I don't know why. Maybe worth checking with WMD

...Sicherlich ^Post 09:42, 19 August 2014 (UTC) by the way; I think here it comes far too much to the impression that de-WP is against changes and needs a lot of convincing. I don't think so. The reall issue was and is the superprotection used by WMF. IMO that was panic. There was no need to act rashly. Probably the MV would have been back online shortly after anyways and the whole discussion would not have been occured.

• Who should be a part of the communication distribution?

WMF wants to communicate so WMF has to do it? But you should look for key persons to tell you where and what is going on. best would be to be part of the community but it seems the WMF is a body who has little to do with the people actually running e.g. Wikipedia. Maybe thats a problem too.

• Why are personal insults a part of our methods of communication? How do we stop that?

a question at least as old as Wikipedia. A lot of solutions were proposed. None of them worked so far. I don't have a new one

• How do we stop pointing fingers?

do we have to? I would see that as taking responsibility.

Planning for the future

• Is it possible for us as a community as a whole … to think beyond what we do today

do we need to? If someone (community, WMF, whoever) comes up with an idea; "sell it" - if it is a cange for the whole community; convince them that it is a good idea. If it is only a small group who will be effected or it's a project with rather no effect on people who are not interested; discuss it only with the interested ones

Comment - Thank you for these thoughts. With Tech News and other methods of communication people can find the information even if they aren't interested, even if it doesn't affect them directly. That being said, they may be interested in learning about it. _Rdicerb (WMF) (talk) 17:23, 18 August 2014 (UTC)

The issue with tech news are the tech news. I had them on my watch list for several years. But I'm a Wikipedian not a technician. Most of the things there I dont get: "Add mw-ui-checkbox", "Add mw-ui-* context colors forany element", "jquery.textSelection: Remove references " ... WTF?? :o) - That is not selling the idea. for me it's just kind of gibberish. And even if I get it; if McDonalds or apple would sell like that they would already be bancrupt :) ...Sicherlich ^Post 02:03, 19 August 2014 (UTC)

• WMF is a technology org. Do you disagree with that mandate?

No. IMO correct. Maybe some minor services beside.

• What did you like that WMF development has done?

Single user login

I hope there are more who just don't come to mind. …

VisualEditor; great idea for new Wikipedians. but there is too much room for improvement to say its a good one

vector: a lot of money; hardly an effect (moved the search box from left to right; as a reasearch done by asking 8 (yeah, eight) people.

wikidata: good idea. Nice for interwikis. For more due to the lack of sources for datas not usefull. As well I have no clue how to use datas from there in articles; usability seems to need a huge improvement.

more?

Readers & Editors

• To what degree should we care about the readers?

They are the target group. From WMF point of view it depends: they donate the money you earn. So they are much more important to you than to the Wikipedians. … But if you annoy of the Wikipedians they will simply leave. as a long time result you'll have less readers and less people to donate. But that will take effect in years or decades. So the current WMF employees will not feel the effect

• Should we split the site into reading and editing experience? New users and contemporary?

Reading and editing; maybe. I already did it for myself. No Mediaviewer as useless if you want to use pictures on wikipedia (no filename shown). No visual editor as not usefull and monobook as I like it more.

New users and contemporary: please no.

• Do we believe that the expectations

maybe but what are UX studies?

• Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?

No, not at all. The readers are the goal.

Features

• Do you think old features need to be redone or removed

why is it needed?

• What are the top 3 product features … for readers

don't know. Maybe a better search enging? But they as far as I know use google/yahoo/bing… anyways :) …

• What are the top 3 product features ... editors

wikidata - with sources and usable for non-IT-Nerds. A lot of work is done several times in different languages: e.g. populations of cities; you have to change it in every language. Waste of the time of people we are lacking; editors. But to use it you need the possibility to give sources and someone "normale" who is not an IT-freak (e.g. me :D ) should be able to use these datas and if necessary change them. As on WD is already a lot of datas without sources its probably getting hard to convince the communities (at least on de) to accept Wikidata. Maybe the rollout was done to hastily.

WMF

• How can we avoid WMF

see WMF as a service body for the editors. The current behavior seems more the other way around. (Seen with the implementation and use of Superpowers. In a computer game I think it's called god mode .oO )

• By which change of structure/communication

Same answer: "See WMF as service body". If you think you need some change to be done: sell the idea. Instead of using the power of technique use the power of selling/convincing

• How do we avoid … regularly leaving

see the 2 answers before. ..

• How do we manage the aim of project-wide votings .. into the languages of the communities

Probably it would help. I thought that was on of the reasons for communtiy advocates. (But looking at de the idea might be completely different. Maybe the word community is misleading?)

• How can we roll back the influence of financial and marketing aspects on decisions …

Sadly I think we can not. WMF staff is working for money. WMF has the power. Why in gods name should they make decisions which gets them less money? That would be stupid. Its an Principal–agent problem and the community is not in the position to change e.g. the ways of payment or distribution (there are some minor possibilities; but they are IMO rather limited to keep the people quiet and satisfied without a relevant impact for the WMF as a whole)

...Sicherlich ^Post 10:38, 18 August 2014 (UTC)

Answers by TheDJ

Decision-making

• How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?
  • Emphasize the stupidity of the status quo in light of the advantages of the way forward. It's the only thing that can work. Those advantages have to be actually there though...

• How can we reach representative decisions for all wikis? Good practices in product tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?
  • I agree, but when doing this, we need to recognize that before vs afters are disruptive. We need to make informed decisions and onboard people before. evaluate, backtrack or improve after... A product is not done when it's 'funding' ends. That's usually when it actually starts and requires the most support. This is often overlooked.

• Do we think RfCs to be the best means of achieving a decision or consensus? -- They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?
  • We probably have nothing better. I would say that these tend to work better if they have experienced people guiding them to make sure the best information is collected and the best decisions can be made. We could use better software to evaluate and comment on designs, and probably we need better software to process feedback (phabricator might be able to assist in both of those, but it's probably not a holy grail).

• Does every decision need to be community vetted? If not, where is the threshold?
  • No, that would be terribly inefficient. In my opinion these things only go wrong when WMF loses sight of basic community principles (example, launch a mediaviewer presenting missing and broken attribution for images, before guiding the community on how to make license/attribution information machine readable. Reasoning from the principles of the community it is clear that the structured data project for files would have to be done before the actual launch of MMV).

• Does every decision need to be WMF vetted? If not, where is the threshold?
  • Definitely not, the community usually self organizes just fine. WMF should intervene only when needed. Legal, long term stability, short term stability etc. TheDJ (talk) 11:01, 18 August 2014 (UTC)

Communications

• Where should we communicate to reach enough critical people so we don’t get into a place where we waste work? Are we using the right communication channels to reach contributors? Readers? How do we prevent ourselves from jumping in at the very last moment (if we missed getting engaged on-time) when the works is ready and public?

• • Inform on wiki, heavily invest on flexibility of receiving information. Everyone should be able to browse the broadcasted information, but you only get personal notifications for stuff that interests you (Do you want more tech news: yes/no, do you want to test new version of the software: yes/no, do you want more news notifications on community governance: yes/no, use lots of metrics here). Echo has been quite popular, we should invest a lot more there, it should be our primary communication and action framework. Feedback should be centralized, integrate translate links to assist people with foreign languages, thumbs up and thumbs down voting of comments might be useful here. Reception of feedback should be acknowledged, people like being acknowledged.

• Who should be a part of the communication distribution? Should it be exclusively the Community, the WMF, or a combination of the two? If a combination of the two, how should responsibilities be broken down and shared?

• Why are personal insults a part of our methods of communication? How do we stop that?
  • Insults are the easiest way to show you are serious in an online world and to spur people into action.

• How do we stop pointing fingers? How do we align ownership, execution and authority without blocking all changes?
  • We don't, we simply need to be so good, that pointing fingers is not required...

Planning for the future

• Is it possible for us as a community as a whole (and as many smaller communities) to think beyond what we do today: editing long-form Wikipedia articles? If so, how would we do this? I'm not saying this because I don't think we should work on making long-form editing easier, I want us to recognize that there are more types of content and ways to contribute and consume it today.
  • For the movement yes. For Wikipedia editors.. eh no. I don't really think so anymore. What I have seen from people over the past 2 years, they are very stuck in the long-form of Wikipedia that we all love and hate.

• What external threats should we be planning for?
  • Commercial entities being better at what we do than we are, but not caring about the moral reasons of what we do.

• Practices in Technology
  • Document the morals/principles/sensitivities of the communities. Review designs and 'products' on those issues. Take on more technical debt, focus less on products and more on workflows. Many of of products are failing on being too complex without fixing smaller things that they depend on, creating confusing and inflexibility. Fixing 'small stuff' will give use more flexibility.
  • Another point here. Our design team is VERY young. They are doing great work, but they need to still find a way to collaborate with community on this. This is a completely new area. The design they do is very dependent on the workflows that need to be supported, and they need to find better ways to be open on critique on their analysis of these workflows. We need to teach and find ways for them to learn this.

• WMF is a technology org. Do you disagree with that mandate?
  • WMF is an extension of the community first. It does stuff that individuals cannot do quickly/reliably enough. Software is one of those things. Legal another, servers, administrative assistance, etc

• Our current change-management practices around Site Operations/DevOps are dangerously out of date (especially how we change production code). How should we fix that?
  • I don't think that is something the community can answer. Says who ? perhaps we are just different ? Perhaps being different is good ?

• What did you like that WMF development has done? What parts of the process worked? Which parts of the process did not work specifically, and why?
  • Mobile, Echo, cirrus search, Lua/Scribunto, mediawiki.ui, the JS deprecations, parsoid, citoid, mathoid, PDF renderer. Total failures: Original UploadWizard, TimedMediaHandler, ArticleFeedback, Mobile uploads. Things that usually tend to work are things that are less 'in your face, less confrontational'. Things that tend to not work, are stuff that is not solving 'underlying issues', but just builds pretty UI on top of crap, stuff that miscommunicates community principles, stuff that creates more editor work without having an editor workflow.

Readers & Editors

• To what degree should we care about the readers? Reach?
  • A lot: better question is, what are we doing for editors and is investing in editors not a better way to get to the readers ? In basic experience, how are we not serving the reader ? We are not telling them our attribution information on images, we are not helping them deal with referenced works, we push pages that are 'too big' for their device, we overload them with 250MB images. These are the questions we need to ask, not "What kind of product can we make for the reader". Too often we take this 'product approach'.

• Should we split the site into reading and editing experience? New users and contemporary?
  • We should stimulate people to gradually 'learn' the site and get access to different and more advanced tools. Don't overload, don't 'strip', but suggest things and dock them into their interface once they are used to them. (app’ify the workflows and how they are discoverable).

• Do we believe that the expectations of both readers and contributors have changed over the last 10 years? Our UX studies show this to be the case: the users who come to Wikipedia today get confused because they have been trained to expect different UX patterns. How should we handle this?
  • Gradually iterate, but get the basics fixed first. So build that common ui and style kit, and encourage editors to use those building blocks for their tools. Fix the presence of machine readable image information. Create the databases for curating edit steps (is this a good picture to illustrate this article yes/no). The 'different' UI is too often dependent on other information, that we have not build properly enough or not at all. Our landscape is too complicated to build interfaces that work for 1 workflow of the 98% and the 10000 workflows of the 2%. We need these basic building blocks to make sure that the community is able to come along in that transition. To rebuild their own tools in better and more sustainable ways. We need to show the advantage, lead by example.

• Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?
  • Course not and it isn't only about readers either. Or only about Wikipedians. But there is a very large, longstanding and important link between the wikipedia editor community and the mission, and it would be arrogant to ignore that.

Features

• Do you think old features need to be redone or removed? What would you do to clean up old cruft and features? How would you tell the few hundred editors that would get really upset?
  • Everything media. We are 10 years behind there, yet it is one of our biggest assets. Rebuilding it will be tough. But need to start from the ground up. We started sky down (MMV)... Why for instance, is the database structure behind the files still not fixed (as a WMF contractor mentioned to me at WikiMania) ? That is an important infrastructural element complicating the Files management software that we are not addressing, these kinds of questions are important but far less visible than 'Products'. Is that why we are not doing them ?

• What are the top 3 product features that you would like to see rolled out over the next two years for readers, assuming we could meet your quality standards?
  • VE, structured data for Files, Maps correction: The Getting Started-technology is much more important and should be used for more than getting started for readers.

• What are the top 3 product features that you would like to see rolled out over the next two years for editors, assuming we could meet your quality standards?
  • Echo improvements, Mathoid, structured data for files.

@TheDJ:, as usual, I find your thoughts compelling and I agree almost 100%. The only points on which I disagree are on the first and last questions of "planning for the future." I strongly disagree with the widely held notion that Wikipedians are averse to change; rather, identifying, addressing, and "selling" changes that satisfy a highly varied constituency are all challenging tasks, and have not been done effectively, to date, on large software projects. And I disagree on some of the specifics of the last one -- I think there are important problems with Mobile, and I think the rollout of Echo notifications (in general, a good and important improvement to the software) was appallingly and needlessly bad, and we are still dealing with some of the bad feelings that were generated in that fiasco.

There are a few other areas where you have much more experience than me, so I find your answers intriguing, but don't really have a basis for saying I agree or disagree.

But overall, +1 to just about everything. Thanks for articulating all this. -Pete F (talk) 20:54, 18 August 2014 (UTC)

You are right about the interpretation that we need to make the 'product' better and that that would help, but it is about the balance of course. Do we think we can achieve what so many sites cannot achieve ? Not even Facebook and Google have product transitions that never end up in a revolt... Are we so confident in our 'short term' development capabilities, that we are willing to put such high requirements on our software ? Does that help the long term progress of our websites ?

I do think these communities are averse to change, but not more than much more than the average population that is committed to, invested in and/or proud of their hobby. The aversion is logical, you try to standardize and create routine in order to deal with the complexity. But it is also dangerous, because you stop considering. You stick to ways, not because they are good, but because they are familiar, unable to consider why they exist in the first place. Wikipedians/Commoners stick to ways, like none of our other communities even get close to, it's not a bad thing, but it's definitely not a good thing either. External influences need to (and will) shake that up once in a while. If it's not the WMF, it will be people even less close to us.

With regard to Echo.... It's an iPhone with an antenna that requires a bumper.. But i'd rather have an iPhone than a drum ... :D TheDJ (talk) 22:11, 18 August 2014 (UTC)

"Not even Google or Facebook..." -- Although the financial resources of these companies dwarf WMF's, we must not forget that we have a one tremendous advantage over them: a community that only ever came together because of its alignment around a single vision. Google is trying to develop software that maximizes its corporate value, and user satisfaction and retention are mere variables among many others in a formula defined by stock price (and, if you're feeling generous, "non-evilness"). But Wikimedia is fundamentally different: we are all here to serve end-users (readers). So it's a different problem, and -- I believe -- a simpler one. If we can have coherent and reliable communication, we have the ability to achieve general (but never perfect) consensus. So perhaps we should aim to exceed what Google and Facebook can do, rather than justifying shortcomings with a sentiment like "if the big boys can't do it, neither can we."

Yes, change-aversion exists, but so does hunger of change. I do not say that change aversion doesn't exist, but rather that an over-emphasis on this phenomenon is a cancer at the WMF, and has started to become a self-fulfilling prophesy. A cycle that looks something like this:

1. We want to accomplish XYZ
2. The community will never accept XYZ
3. The community will never accept anything
4. We will consult the community because we must; but we expect them to reject XYZ
5. That rejection is meaningless
6. XYZ! with a few minor tweaks to show that consultation was heeded

On Echo -- I absolutely agree with you that it is fundamentally better than the system it replaced, and I think most of the English Wikipedians who were offended by the way it was pushed through did as well. In fact, the same may be true of Media Viewer -- I certainly think that something that looks and acts more or less like MV could be a great improvement over the current file pages. But when you take good software (like Echo), or even potentially good software (MV) and push it forward in a way that sparks (expensive, if none of the other problems bother you) outrage and protest, perhaps that should be a sign that resources are being invested in the wrong kind of activities. -Pete F (talk) 22:51, 18 August 2014 (UTC)

Things that need to be free

To the question on what the Wikimedia projects are doing: no, there is a lot more to do. (WMF focuses on some small corners of what we do, and we'd need to do much more.) The Wikimedia projects should have several orders of magnitude more content and hundreds millions editors. --Nemo 22:37, 18 August 2014 (UTC)

Answers from Kww

Decision-making

• How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?

• Subordinate. Remember that the WMF exists to serve the editing community, not the other way around. Each version of Wikipedia has its own method of reaching decisions. WMF's responsibility is to choose the priority of supporting those decisions, not to make them.Kww (talk) 05:34, 19 August 2014 (UTC)

• How can we reach representative decisions for all wikis? Good practices in product tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?

• One wiki at a time. There's no method that works for all wikis, because we are all independent of one another. Kww (talk) 05:34, 19 August 2014 (UTC)

• Do we think RfCs to be the best means of achieving a decision or consensus? -- They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?

• They are the only method that English Wikipedia has gotten to work for it. Other versions have different methods.Kww (talk) 05:34, 19 August 2014 (UTC)

• Does every decision need to be community vetted? If not, where is the threshold?

• Every decision without legal ramifications can be community vetoed.Kww (talk) 05:34, 19 August 2014 (UTC)

• Does every decision need to be WMF vetted? If not, where is the threshold?

• For legality, yes. Otherwise, you are the servant, not the master, and your opinion doesn't matter very much.Kww (talk) 05:34, 19 August 2014 (UTC)

Communications

• Where should we communicate to reach enough critical people so we don’t get into a place where we waste work? Are we using the right communication channels to reach contributors? Readers? How do we prevent ourselves from jumping in at the very last moment (if we missed getting engaged on-time) when the works is ready and public?

• By not making things default until there's a consensus to do so. Every feature you make is a beta test, and if you do it well, people will enable it eventually.Kww (talk) 05:34, 19 August 2014 (UTC)

• Who should be a part of the communication distribution? Should it be exclusively the Community, the WMF, or a combination of the two? If a combination of the two, how should responsibilities be broken down and shared?

• I don't quite grasp what you are asking.Kww (talk) 05:34, 19 August 2014 (UTC)

• Why are personal insults a part of our methods of communication? How do we stop that?

• Because the WMF plays an annoying game of public relations, making demonstrably false statements, denying obvious truths, and intentionally misleading people. Eventually, we get tired of being lied to and snap back. As I said above: promise us that the next time a staffer tells a bald-faced lie he will be fired immediately and you will see community relationships improve substantially. Unless, of course, they lie again and you fail to follow through on your promise.Kww (talk) 05:34, 19 August 2014 (UTC)

• How do we stop pointing fingers? How do we align ownership, execution and authority without blocking all changes?

• By remembering that you are the servants, not the masters. If you have made the communities you serve unhappy, you need to rectify that, not try to tell those communities that you know better.Kww (talk) 05:34, 19 August 2014 (UTC)

Planning for the future

• Is it possible for us as a community as a whole (and as many smaller communities) to think beyond what we do today: editing long-form Wikipedia articles? If so, how would we do this? I'm not saying this because I don't think we should work on making long-form editing easier, I want us to recognise that there are more types of content and ways to contribute and consume it today.

• Given that we are an encyclopedia, I don't think that we should plan for not being an encyclopedia. There are all kinds of different wikis that can absorb, rehost, and manipulate our content to their heart's content.Kww (talk) 05:34, 19 August 2014 (UTC)

• Practices in Technology

• WMF is a technology org. Do you disagree with that mandate?

• Most emphatically. Your mandate is to host servers for the various Wikipedia communities and maintain software.Kww (talk) 05:34, 19 August 2014 (UTC)

• Our current change-management practices around Site Operations/DevOps are dangerously out of date (especially how we change production code). How should we fix that?

• Primarily by not making changes in the default software until it works. Wikipedia is too large of a project for Agile-based methodologies to be worth consideration for software that is enabled by default.Kww (talk) 05:34, 19 August 2014 (UTC)

• What did you like that WMF development has done? What parts of the process worked? Which parts of the process did not work specifically, and why?

• The maintenance of the existing code seems to work reasonably well. Lua is the only new development that comes to mind that was both useful and well-done.Kww (talk) 05:34, 19 August 2014 (UTC)

Readers & Editors

• To what degree should we care about the readers? Reach?

• Worrying about the readers is each community's responsibility, not the WMF's.Kww (talk) 05:34, 19 August 2014 (UTC)

• Should we split the site into reading and editing experience? New users and contemporary?

• No.Kww (talk) 05:34, 19 August 2014 (UTC)

• Do we believe that the expectations of both readers and contributors have changed over the last 10 years? Our UX studies show this to be the case: the users who come to Wikipedia today get confused because they have been trained to expect different UX patterns. How should we handle this?

• By allowing us to absorb and train them without getting in the way. Providing us with tools that are easier for novices to use but don't allow novices to see what needs to be done to finish the job makes it more difficult to deal with them in the long run.Kww (talk) 05:34, 19 August 2014 (UTC)

• Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?

• Your mission is about the editors. Ours is about the readers. When we need help doing things for readers, we ask for that help.Kww (talk) 05:34, 19 August 2014 (UTC)

Features

• Do you think old features need to be redone or removed? What would you do to clean up old cruft and features? How would you tell the few hundred editors that would get really upset?

• We don't have very many "old features" that actually interfere with our mission. Those "few hundred editors that complain" are very likely right when they tell you not to change it. They are the ones that actually produce the content that people come here to read. It can get irritating. For example, here I am, still stuck with Echo stripping away my talk page notifications. It can't even give me a diff of my last talk page message, instead making me search on my history tab and find the diff link myself. It added some nice features (like being notified of rollbacks and mentions) but took away important old ones, and the developer community only heard "old editors don't like it" and never cared to read our legitimate complaints about the changes.Kww (talk) 05:34, 19 August 2014 (UTC)

• What are the top 3 product features that you would like to see rolled out over the next two years for readers, assuming we could meet your quality standards?

• I think the current reading interface is nearly perfect for desktop devices. The mobile platform needs substantial work, but I'm not the one to tell you what needs to be done there. All I know is that the first thing I do when I use Wikipedia on my phone is to switch it to the desktop version, and I curse whenever the mobile UI comes back.Kww (talk) 05:34, 19 August 2014 (UTC)

• What are the top 3 product features that you would like to see rolled out over the next two years for editors, assuming we could meet your quality standards?

1. An all-out focus on a protection and checkuser mechanism that makes it genuinely difficult to create and use sockpuppet accounts. Our current mechanisms are laughably weak.
2. The ability to protect individual articles against IP ranges and individual editors. Most of the time we resort to semi-protection and full-protection, it's to prevent one single editor from being a problem. The more we can target protection, the less collateral damage protection we will have.
3. A revamped Visual Editor that focused on the actual problem: assembling templates, wikitext, HTML, and other components together into an article. VE used a Wordpress editor as a model when it should have been thinking of Scribus.Kww (talk) 05:34, 19 August 2014 (UTC)

A few comments and additions by ca$e

I concur with Kww on 99% of each and every question. Thus, only a few minor additions or slight derivations:

Regarding RfC: As Kww already pointed out, you should not assume that the decision processes of enWP are the same in other languages and projects. A notable example is deWP: decisions there have, in the case of a "Meinungsbild" (nowadays roughly: poll for decisive guidelines for future process!), a much more binding effect than RfCs in enWP, which more resemble a "Umfrage" (roughly: poll for perspectives!). Especially, every admin strictly has to follow the outcomes of the former. Thus, it was seen as a matter of severe ignorance when WMF employees intervened against it. The blocking of Erik's account for such a grave breach of deWP's decisive rules was the necessary outcome, as everyone would have been able to foretell. The same would and will hold for every other breaching of those rules. (We already told you of these differences and circumstances, while the phrasing of your question ignored them again.)

Also, regularly people voice they perceive a very large difference between deWP and developments in enWP / on WMF's behalf regarding focus on quality content. I freely translate e.g. from User:Magiers:

What WMF is seemingly working towards is more foundational. It is not merely about a better presentation for readers. It is also about making WP more open for everyone. Everyone shall be able to contribute every kind of content. This is diametrically opposed to the efforts towards quality especially in deWP. Our criteria for relevance, our efforts to delete unwanted trivia from articles, our insistence on quality sources -- all this in effect makes it impossible that everyone can participate in every way. For example, when people on meta dream about showing videos of tigers in articles -- i can visibly imagine how (not) very excited our editorial project for biology would get, when people would submit a bunch of mobile clips. Indeed, we might currently scare off some people because of tech issues. But if we would drastically simplify these -- which would, of course, be desirable -- we would still scare off some people by not accepting their contributions because of their content (...). In the end, the main conflict with WMF will surface: Do we want to work towards a reputable encyclopedia, for which often significant research is needed to contribute (...), or just another platform for everyone to contribute anything.

Regarding readers and tech and "top product features" etc: You should realize that very many (myself e.g. included, in cases where i really only want to read - in every other case, i, like Kww, curse the mobile version and switch to the desktop view!) often read WP by other means: For mobile phone and tablet users, there are apps that provide a very convenient layout (one popular example for deWP would be http://dasreferenz.com/), or web services (like e.g. http://www.wikiwand.com), or readers such as getpocket.com for longer articles. I also find the idea mistaken to invest larger resources in developing an environment to edit WP from phones or tablets. The time has long passed that we would have profited from stubs of, say, the length of a twitter post. We also do not at all need millions of useless pictures, and especially not without proper licensing. What we need, at the moment and in the coming future, are more well-versed experts on topics where WP still severely lacks in article quality. For these topics, we need profoundly worked-out articles which people will, probably for several years to come, not write on their phone or tablet, but on their desktop. For those needs, please read and think about e.g. the post by User:H-stt translated here, especially these 2 paragraphs:

What's it like with user recruitment? How many millions have been invested in this over the past five years? Probably a two-digit number. And how many new authors were gained by it? Correct: practically zero. Why? Like Micha above I don't see the problem with the editor. Anyone who has the intrinsic motivation to contribute to the greatest free education project in human history will not be deterred by the editor. There are many other barriers that are more important (first and foremost the ability to prepare information appropriately, but I mention this only as an aside). So here, too, the Foundation mucks about, but has achieved exactly no demonstrable results.

Why is it all like that? Because the Foundation (and in a similar way but to a lesser extent also WMDE) has grown much too fast. The unlimited money supply from the fundraising campaigns shows the tremendous enthusiasm of our readers, but it has seduced people into hiring staff without first agreeing on goals and methods. This excessive staff and bureaucracy then very quickly became estranged from its base, the community, and is now fighting for self-preservation. As far as content is concerned, they have nothing to show, so they have to use force.

Further, regarding top product features: I agree with Llywrch below, that, e.g., the "thank you"-notices were one step in the right direction, and also the evaluation: "I think it's important to note that they are all small innovations that proved easy to understand & adopt -- or not to use & ignore." However, i would also like to see mechanisms implemented that would amount to large innovations. Among them, better functions for citing and referencing (please evaluate this in the context described above regarding the focus on quality development in article space). As you can probably already deduce from the overall perspective outlined above, among the most important projects i see e.g. https://en.wikipedia.org/wiki/Wikipedia:The_Wikipedia_Library , projects that directly fund quality content contributions, as well as e.g. https://www.mediawiki.org/wiki/Extension:Proofread_Page for WikiSource. We also have had extensive collections of input regarding lacking and much needed features, lastly especially https://de.wikipedia.org/wiki/Wikipedia:Umfragen/Technische_W%C3%BCnsche . In the meanwhile, however, we have lost key functions by changed circumstances for the tool server. My topmost priority for WMF would be: let your developers repair and again provide those key functions -- or rather, with one provision: At the moment, we also have lost several admins and bot operators (e.g. of the for deWP much needed GiftBot) in protest against WMF's aggressive behavior. There are therefore currently very many important maintenance jobs undone because of you; thus, you should immediately focus on repairing recent aberrations of WMF employees and proper amends towards affronted communities! Afterwards, i would indeed advise you to use means such as just pointed to, in order to evaluate what editors -- and by the same token (see Kww!), readers -- really want and need (and notably before wasting more resources by working in undesired directions!). This would amount to a more representative survey than several methodologically ridiculous pseudo-surveys WMF undertook recently and also more than a mere opinion poll on your user page. However, judging from the answers you already got, not few of them from some of the best experts regarding UX like Pete we have here, one could already classify this poll as a successful semi-structured expert interview! At least, i appreciate this as a previously too much neglected step in the right direction. However, what counts in the end is visible output... Ca$e (talk) 08:13, 19 August 2014 (UTC)

Your inheritance (Another long rant)

Hi. This isn't exactly related to the current issue but more of a long term problem. At the risk of sounding a little blunt here, I'll tell you something that no one is going to tell you honestly - You have inherited a lot of messes. It's unfortunate and unfair, but you are not starting with a clean slate. There is a lot of history that had nothing to do with you. Sue was an awesome leader and built WMF from the ground up, but the same people that she trusted to oversee things are still taking the lead - the rest of the staff (including you), the board and everyone else on WMF's side - just does damage control and chooses to stand united, taking the same positions about neutrality and cooperation - You will see this pattern a lot where you'll just be catching up to other's decisions. This seems like an attempt at damage control here, and really reading pages and pages about this will not have an effect unless you take a stand. You'll realize more and more that before you even approach any issue with a wikimedia community - they would have an opinion about you, there would be a lot of distrust between the two sides because of the shared history.

This is not the first time de.wp has had such divergent positions - I remember image filter had a particularly strong reaction from de.wp community, then there was that issue with it.wp, regular issues with commons, controversial content, Chapters and fundraising issues were particularly bitter, all the while every other project(we have hundreds) feels they are the black sheep of the family and en.wp is all that gets the focus of WMF attention because it's the flagship. All that is not to say en.wp hasn't had a lot of these issues at regular interval - probably more so than others, but you can see a pattern if you research this. Similar problem affects the chapters, editors, arbcomms and any influential entity you will encounter - they all need respect and attention. You know that forcing things on people they don't want as a form of tough medicine doesn't work with open communities - open source communities love forking. If what they like about this gets changed too much, they will just leave and start fresh with a recent backup. And you will see threats of forking as the ultimate deterrent the community has. One of these days, at the right confluence of time and opportunity, WMF just might have to deal with its first fork in the road. You should weigh what matters more to you before that happens.

Erik's recent argument was that WMF isn't the type of organization that reactively supports the community with small requests and does things the community wants (I'm not sure if he was proud of saying this - he shouldn't be) - mostly because he has hired close to a hundred new staff in the last year or two and they need something to do. Even if it means they design sub-par products that are not requested, that end up wasting millions, that eventually have to be forced upon people who never wanted it in the first place, so either they end up in the dust bin or they are tolerated. It might entirely be true that the staff numbers has gone up twice and the relevant work the community wants, has gotten less instead. The only joke is, community had nothing to do with this expansion, erik et al decided it was time to expand aggressively and build a few things, basically now that there are a few people with hammers - they need some things to break. None of this expansion has achieved WMF's larger goals - the editor decline hasn't reversed, the community is not a better or more diverse place than it was a couple of years ago, the relation between the foundation and community are bad as ever, while trust and communication is strangely worse now than it was before. At the core, you know what the issue is and what you need to do to make this go away. Respect the community's wishes - They built it, they will use it - they should have more of a say on it than you or erik or anyone else. You want to address the larger problems and find a new process for development/communication/decision-making - that is excellent, but those discussions can not happen while you are standing on someone's toes - de.wp need to have their concerns addressed and things return to normal, before a new process is found to work with the community or they will always remember this - and it would be another mess you inherit, except you could have fixed it here and now. Thanks for reading. Theo10011 (talk) 08:15, 17 August 2014 (UTC)

Some answers from Pete Forsyth

Lila’s questions:

Lila said: (trimmed to highlight the parts that seem most relevant to me)

it is top of my mind.

We clearly opened up a really big issue that needs resolution.

I don’t want us to rush.

I want to understand this and think it through and to work together to make this work.

We need to figure out how.

^^^ All of this is excellent news, I agree, I am happy to hear it, and thank you for stating it explicitly.

I will answer some of these questions below:

Decision-making

• How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?
  • The most important thing is to have real discussions about what "forwards" vs. "backwards" looks like early in the planning stages, and consistently throughout, major changes. True, you will never reach 100% agreement, so 100% agreement cannot be the standard. I include the word "real" intentionally -- I think with recent software changes, there have been efforts to do this, I believe the software teams have all operated with a good faith intention of adequately surfacing and dealing with those issues. But dealing with them is a difficult art, and I think in spite of the good intentions, there has been rather consistent failure to do this effectively.

• Do we think RfCs to be the best means of achieving a decision or consensus?
  • In my opinion, absolutely not. An RfC calling for the removal of a major feature is something that only happens when many things have been done wrong leading up to deployment. It's better to work toward a reality where RfCs on major topics like this are not needed to begin with, than to try to figure out a way to do RfCs on these topics better.

• Does every decision need to be community vetted? If not, where is the threshold?
  • See above. IMO, all decisions should be vetted in proportion to how sweeping the changes will be; but that vetting should be incremental, and should involve sampling, A/B testing, etc. It should be done in a sophisticated way, not merely posting notifications and answering questions. There's no quick answer to what a sophisticated way looks like; it will take a lot of work to design something that works well for Wikimedia. But the idea that it is impossible to do, I do not think is compatible with the purpose of the WMF.

Communications

• Where should we communicate to reach enough critical people so we don’t get into a place where we waste work? Are we using the right communication channels to reach contributors? Readers? How do we prevent ourselves from jumping in at the very last moment (if we missed getting engaged on-time) when the works is ready and public?
  • IMO this has been a bit of a red herring in the current discussions. It's an important question, but if processes are handled well, the announcement process will naturally be a lot easier, because you will have lots of volunteers and stakeholders helping you find the right ways to spread the word. In the case of MV, I had very little willingness to engage deeply in the process -- and that is true of many other Wikimedians with much to offer to something like this. I'd advise looking more closely at how these processes are conducted than at how they are advertised.

• Why are personal insults a part of our methods of communication? How do we stop that?
  • An excellent, excellent question. I think you are in a uniquely powerful position to interrupt that dynamic by insisting that WMF staff not participate in "insult culture." It is not a magic bullet, but it will have a big impact.

• How do we stop pointing fingers?
  • By relentless focus on the bigger goals and aspirations that unite us. And by making the effort to clearly document and learn from the more successful efforts to institute broad change.

Planning for the future

• Is it possible for us as a community as a whole (and as many smaller communities) to think beyond what we do today: editing long-form Wikipedia articles? If so, how would we do this? I'm not saying this because I don't think we should work on making long-form editing easier, I want us to recognise that there are more types of content and ways to contribute and consume it today.
  • I look forward to seeing where this question leads you. One of the things I find most inspiring about the Wikimedia community is that this sort of thing pervades. But I don't think we've found good ways to capture and organize those kinds of discussions effectively. Much work to do.

• What external threats should we be planning for?
  • We should assume that Google, and probably other big companies, have active plans to fill the void if Wikipedia should falter. That's not to say they are actively seeking to undermine it. But in my opinion, the biggest threats are internal, not external -- and there are several kinds.

• WMF is a technology org. Do you disagree with that mandate?
  • Yes, but then again, it's sort of meaningless to agree or disagree unless it's fleshed out a little more clearly what's meant by that.

• What did you like that WMF development has done? What parts of the process worked? Which parts of the process did not work specifically, and why?
  • Lua. It seems to have been carefully planned out, communicated effectively to stakeholders, rolled out in a way that didn't cause disruption, and communicated effectively in a general-purpose blog post.
  • Perhaps Vector -- I was not intimately involved with that one, but it seemed to work fairly well.
  • The Upload Wizard's initial rollout.
  • I should note, I worked for WMF during the last 2, and it's entirely possible my perspective was heavily skewed by that.

Readers & Editors

• To what degree should we care about the readers? Reach?
  • I can't think, off the top of my head, of any reason anybody would be here if they didn't care about readers and reach. Can you explain why this question seems important? I think we all care deeply about readers -- I can't really think of anything we would agree on more. It's why we're here.

• Should we split the site into reading and editing experience? New users and contemporary?
  • A bold and interesting idea. I'd love to see it spelled out and opened for broad and genuine discussion, independent of a timeline for implementing it.

• Do we believe that the expectations of both readers and contributors have changed over the last 10 years? Our UX studies show this to be the case: the users who come to Wikipedia today get confused because they have been trained to expect different UX patterns.
  • Yes, of course. The strong belief at WMF that the core editing community is change-averse is entirely unjustified. It's just that the specific changes, and the specific processes for implementing them, have been deeply flawed for a long time.

• Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?
  • I'd be fascinated meet anybody who says "yes." I suppose there are a few people who are simply dedicated to updating some little corner of Wikipedia, like making sure all football statistics are kept accurate and up to date, out of a personal obsession. I've encountered some of those people, and they are tremendously productive, and sometimes a bit irritating to deal with if you need them to step out of their narrow view and consider something from a different perspective. But most of all, I almost never hear from this sort of Wikipedian in broad discussions. So I can't imagine in what venue they would answer "yes" even if that's what they feel in their hearts. To put it simply, they are not socially influential.

Features

• Do you think old features need to be redone or removed? What would you do to clean up old cruft and features? How would you tell the few hundred editors that would get really upset?
  • Absolutely, and Wikimedia projects are full of suggestions of how this can and should be done. We have these discussions all the time. Sometimes they go well, sometimes they are abysmal failures. There is a great deal to be learned from looking at how on-wiki discussions go.

• What are the top 3 product features that you would like to see rolled out over the next two years for readers, assuming we could meet your quality standards?
  • OK, I'll take up this question merely as my personal view. I want to be clear -- this is a good faith effort to engage with the question as asked, and does not reflect a carefully considered view! I'm sure if I spent more time thinking this through, my "top 3" would change a lot. But Lila, I think you're looking for "what's on your mind right now" -type answers -- yes? That's how I'm taking the question.
  • A good Media Viewer would be fairly nice, but there's no way it would make my top 3.
    1. A collection of features that fix the routine interruptions to discussion you get when somebody links to a discussion, and then the discussion gets archived. The link breaks. You might not consider this a "reader feature" and that's a legitimate point -- but there are readers who do take a great interest in discussions, without ever participating in them. I will include in this bullet point something else a bit debatable: better history and diff-viewing features, that make it easier to find the change you're looking for. It's possible that Flow addresses some of this stuff, I really don't know -- but I'm worried that it will break a whole lot more than it fixes.
    2. Along the lines of the things you discussed in the Guardian interview, there's so much more that can be done to automate and augment routine tasks. I think the greatest potential here is in the volunteer community, because there's much deeper expertise in how things work. So, I don't know if you'll consider this a legitimate answer either, but I'd say WMF could improve the reader experience by working to improve communication flows among different populations of volunteers -- for instance, article writers and volunteer developers who create bots and gadgets. WMF's role as a developer might be important here, but should be secondary to the things that emerge from the movement.

    3. Example video with captions

       Here's a rather specific one, but I think it could have a big impact: create some software that makes it easy to create and translate caption files. Here's an example: I created the English captions, and I recruited a colleague (new to Wikimedia) to translate them into Portuguese. It was incredibly time-consuming, and I can't imagine doing this stuff on a regular basis unless the tools are better. But if it could be done more easily, it would go a long way toward making free video content machine-readable by search engines, and useful across language barriers.

• What are the top 3 product features that you would like to see rolled out over the next two years for editors, assuming we could meet your quality standards?
  • I'll try to come back to this -- but as you can see from my answer above, I don't see a strong distinction. The main thing editors need is stuff that helps them serve the readers better. So it's sort of a chicken-and-egg question, and it doesn't map well onto an either-or framing.

-Pete F (talk) 19:39, 18 August 2014 (UTC)

Interesting.. on that last feature you mentioned for captions... In my opinion, that one is actually a perfect example of things we should NOT build (another would be video editors...), but should facilitate. As software, it is specialistic, complex and thus hard to maintain. Let others (or encourage others to) build that and let us integrate with it in a maintainable and reliable way (not possible if an external user can't even figure out the license/author/attribution of a file...). Actually, we already had this at some point with Universal subtitles, but it was too early and it basically fell apart because of our immature media apis. TheDJ (talk) 22:37, 18 August 2014 (UTC)

Yes, I agree with you -- I tried to answer this in the spirit of "what are the first things that come into your head." This is something I worked on recently where I saw a big opportunity. But you're right, it would not make sense for WMF to devote resources to "solving" this problem. But facilitating the community thinking strategically about, and solving, this type of problem -- I think that would be a good role for the WMF. -Pete F (talk) 22:54, 18 August 2014 (UTC)

Answers from Llywrch

Since I could write at length in answer to all of your questions, Lila, I'm going to limit myself to only a few of them. And I'll admit at the beginning that my answers are doubtlessly slanted to favor the outlook of someone on the English language Wikipedia.

• Communications: One problem is that unless I devote at least 40 hours to Wikipedia & Meta, I don't have any idea of where the important discussions are. en.wikipedia has a template that points people to important points of communication -- "Centralized discussion". I have no idea if this is implemented on Meta. Doing this would allow activists in the different projects to pass the word along of topics here that could effect them.
  • As an addendum, I'd like to provide a surprising example of a failure of communications. There was a fundraiser for Wikimedia in 2013, as there is every year. Yet I heard nothing about it at the time. I don't mean I missed seeing the banners, I'm saying that there was no announcement that it started or how much money it raised. And while I wasn't interested in reading about it at the time, I assumed it was something the Foundation would like to share with its volunteers. I was curious about how it went, but I found I had to do some serious digging to learn if it did, indeed, happen, & how much money it had raised.
• Challenge for the future. I believe the biggest threat to all of the Wikimedia projects is the possible loss of Net Neutrality in the United States. It is far more serious than any possible new technology or competitor, because it has the potential of restricting general access to the websites.
  • I also object to the characterization of the Foundation as a technology company; this mischaracterization is part of the problem. What the Foundation should be doing is facilitating access of its readers to Wikipedia, as well as access of its contributors to resources. The Wikipedia Library is a vitally important project, & it should be a priority of the Foundation to work for its success.
  • You asked what I like about what the WMF developers have done. I can think of three additions I feel were successes: {{ping}}, popups when the cursor touch the link to a footnote, & the option to thank people for their edits. I think it's important to note that they are all small innovations that proved easy to understand & adopt -- or not to use & ignore.
• Readers & Editors One of the biggest disappointments I have with the Foundation -- although I'm not surprised it has not been pursued -- is research into defining what an Encyclopedia is, & what it should be. This is more important than it might appear: how can we create a great encyclopedia when no one can define what one is? There are books that will teach one how to write a novel, poetry, or a biography, but none that will help someone to write an encyclopedia article. (I know; I've looked.) I've thought about this a lot, & wrote down some of my thoughts elsewhere.

I hope this gives you some sense of the ideas that are circulating at the grass roots, amongst those who are creating the stuff that has made Wikipedia one of the top 10 most visited websites. And I suspect there are a lot more ideas -- probably some better ones -- circulating at the other projects. One simply has to stop paying money to consultants & use the time spent listening to their reports reading the discussions out there. -- Llywrch (talk) 05:48, 19 August 2014 (UTC)

Answers from Ghilt

Decision-making

• How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?

As voluntary interactions are dependent on communication at eye-level, a rejection by a majority of editors should put the introduction of a software feature on hold until the issues are resolved.

• How can we reach representative decisions for all wikis? Good practices in product tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?

Is this necessary, outside of violations of the principles? Most issues can be handled locally.

• Do we think RfCs to be the best means of achieving a decision or consensus? -- They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?

"Democracy is the worst form of government, except for all those other forms that have been tried from time to time." (Winston Churchill) In our case, i would adapt this sentence to 'A quasi-democratic RfC...'. The democratic participation is necessary for a voluntary community-dependent organization. An RfC is better than mandatory voting, which would increase voting participation at cost of voluntarity and motivation, which would probably result in lower Author numbers.

• Does every decision need to be community vetted? If not, where is the threshold?

In regard to software introductions, yes. This lowers general objection in the community.

• Does every decision need to be WMF vetted? If not, where is the threshold?

In regard to violations of our principles or to legal issues, yes.

Communications

• Where should we communicate to reach enough critical people so we don’t get into a place where we waste work? Are we using the right communication channels to reach contributors? Readers? How do we prevent ourselves from jumping in at the very last moment (if we missed getting engaged on-time) when the works is ready and public?

Take a majority rejection by a local community as reason to put the local introduction on hold, until the main issue is resolved.

• Who should be a part of the communication distribution? Should it be exclusively the Community, the WMF, or a combination of the two? If a combination of the two, how should responsibilities be broken down and shared?

The roll-out communication should be done by the WMF. Usually the comments of a community will follow. With an additional user selection process like this, the evolution will be facilitated.

• Why are personal insults a part of our methods of communication? How do we stop that?

I currently see anger on the legally less powerful side and arrogance on the other. Both is understandably human and reduces cooperation.

• How do we stop pointing fingers? How do we align ownership, execution and authority without blocking all changes?

In the german language, pointing fingers is not a big a problem as in the english language (it can be sometimes helpful, but is not the main issue, the problem is), where direct critique without euphemisms is rarely uttered to not appear impolite. Aligning the stakeholders is a result of good communication. Introducing a superprotect without necessity, without prior notice to the communities or without mention in the bylaws is not.

Planning for the future

• Is it possible for us as a community as a whole (and as many smaller communities) to think beyond what we do today: editing long-form Wikipedia articles? If so, how would we do this? I'm not saying this because I don't think we should work on making long-form editing easier, I want us to recognise that there are more types of content and ways to contribute and consume it today.

Well, i usually split a long article of mine into two, if the later lemmata make better sense.

• What external threats should we be planning for?

Use of WP-content without generating clicks, e.g. by Google. Forking by disgruntled communities. Lack of funds for server upkeep.

Practices in Technology

• WMF is a technology org. Do you disagree with that mandate?

Partially, it is a tech org with a commons-based content generation. The commons principle reduces the number of usable communication forms (e.g. no imperative form, no superrights outside of office actions)

• Our current change-management practices around Site Operations/DevOps are dangerously out of date (especially how we change production code). How should we fix that?

To better understand the question, what is dangerously out of date about it? Gradually, by the above mentioned 'on hold until resolved'. Maybe one could increase resolution speed.

• What did you like that WMF development has done? What parts of the process worked? Which parts of the process did not work specifically, and why?

A lot. I even understand the utility of the mediaviewer for readers.

Readers & Editors

• To what degree should we care about the readers? Reach?

As wikipedia generates clicks by providing cost-free content, care primarily about the servers and the content generation and secondarily about the reader experience

• Should we split the site into reading and editing experience? New users and contemporary?

I don't think it to be necessary.

• Do we believe that the expectations of both readers and contributors have changed over the last 10 years? Our UX studies show this to be the case: the users who come to Wikipedia today get confused because they have been trained to expect different UX patterns. How should we handle this?

Naturally. A WYSIWYG editor would help not only facilitate content creation but would also mark for QA more often, which generally results in better readability.

• Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?

No, but primarily. Most readers i ask tell me the layout is ok and some articles are too long and complex. The latter is a matter of article-QA, which is handled by editors.

Features

• Do you think old features need to be redone or removed? What would you do to clean up old cruft and features? How would you tell the few hundred editors that would get really upset?

The third question is a generalization, which i would disagree with. Superprotect was a whole new level.

• What are the top 3 product features that you would like to see rolled out over the next two years for readers, assuming we could meet your quality standards?

A better license depiction in the media viewer, a better mobile version

• What are the top 3 product features that you would like to see rolled out over the next two years for editors, assuming we could meet your quality standards?

If you'll allow four: A WYSIWYG editor to not scare away new editors, easier mathematical formula generation, easier splitting and merging article histories and automatic references by doi, PMID or ISBN.

Additional questions:

• How can we avoid WMF developper staff to insult regularly local communities, why is there an attitude from the paid staff of beeing superior to volunteers work and their decisions? What are the reasons of unpatience and inappropriate communication skills by leading devs?

Better not force update while in dispute

• By which change of structure/communication is it possible, that issues, which are important and vital to certain communities, can be earlier, more respectful and without empty marketing phrases could be adressed by the WMF?

Better communication. Less implementing features ad hoc like superprotect.

• How do we avoid, that longtime productive, engaged, project-loving and decent personalities (like de:Benutzer:Drahreg01 recently and many others these days) are regularly leaving the project because of WMF issues?

Eye-level communication, avoid additional unnecessary inequalities like superprotect

• How do we manage the aim of project-wide votings concerning new developments without banning user from full participation, which are not able to communicate properly in english or are to shy to do so? Would it be helpful to provide official strategies, plans, developments and statements by the WMF with official translation into the languages of the communities (perhaps the 10 biggest WPs)?

Non-legally binding translations into the 10 biggest WP languages with a link to the binding original.

• Is it possible that the WMF board defines, under which terms and conditions (number of participants, right to vote etc) the vote of a community is in their opinion an express of their will (or not) as it is mentioned in the guiding principles stated by Jimmy Wales? (btw these principles are still an inspiration and one of the main reason for volunteers supporting this project over the years?

In an RfC setting one can compare voting participation with previous RfC.

• How can we roll back the influence of financial and marketing aspects on decisions of the WMF, which are increasingly discouraging volunteers?

I would wish for a focus on core competences like providing an environment for voluntary content generation, improving reader experience and, of course, secure server upkeep.

Thanks for reading, --Ghilt (talk) 08:11, 19 August 2014 (UTC)

Location of this discussion

Somewhere above Lila said she's not sure where this discussion should happen. The solution is rather simple, she can move the whole content of this page at Requests for comment/On a scale of billions or similar and continue operating in the same way but in a more "official" setting. --Nemo 07:10, 19 August 2014 (UTC)

Wrong questions asked?

I answered the questions but I wonder if that are not the wrong questions or the wrong issue focused on.

Its focused on how to convince the people to accept changes. And yes, there are always people who don't like a new feature (they would call it bug :D ) and probably they change depending on the feature. - Some like Coca Cola, some pepsi, others drink water :D

The whole discussion was started because of the superprotection. There where 2 (yes, two) reverts before Wikimedia started to act. Why so hasty? Then there was 1 (one) more revert and the superpower was used. Why? Was Wikipedia in immediate danger?

A more relaxed approach and we would have the the MV back online and the whole discussion with a lot of upset people would not have occured.

But WMF is still afraid of loosing control. What would happen if you remove the superprotect?

A: someone would deactivate the MV - I assume Wikipedia would not crash? :D

A1: editwar within the de-WP - probably it would get solved like wheel wars are not new and got solved in the last decade as well. If not: superpower is still there.

A2: no editwar: the de-community would discuss it and maybe wait for the changes they requested. ... so it would be implemented later - harm to Wikipedia? A delay in a new feature; happens all the time to a lot of large companies

B: the MV would not be deactivated as people trust, that WMF would in not so far future change some things so that people accept it.

The current situation shows, that you, WMF, don't trust the community and that you are sure that you are superior to the community. You who where founded to support the community

...Sicherlich ^Post 10:05, 19 August 2014 (UTC)

Some answers from Rich Farmbrough

I am giving examples from MV, not because it is particularly egregious, but because I hope everyone is familiar with the issues surrounding it.

Decision-making

• 1. How do we make decisions that move us forward, rather than backwards, when we cannot all agree? 2. When the easiest choice is the status quo?
  1. First build consensus. Use reasoning and evidence, not appeal to authority and "newness".
     1. Example: In the current MV discussion there are references to a usability study, that indicated problems. Ignoring this evidence leads to a confrontational dynamic.
     2. Example: There are no proper studies I am aware of into how people use images on Wikipedia, and what additional benefits the Mediaviewer will give
  2. Assume Good Faith
     1. Example: With MV, there is no rejection of the software by the community, instead there are quality control issues, and design issues.
  3. The status quo is not always a bad solution. Certainly if broad consensus for change cannot be achieved it has something to recommend it.

• 1. How can we reach representative decisions for all wikis? 2. Good practices in product [development] tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?
  1. This is what Meta was invented for. There are mechanisms for contacting all wikis. Cross-wiki discussions are possible with the completion of a small piece of software called "reasonably efficient interwiki trnasclusion" - or by other technological solutions.
  2. "Desired outcomes" is too narrow. We need to be very aware of side effects.
     1. Example: MV breaking the terms of CC-BY-SA puts WMF in breach of copyright law, and probably destroys its shelter under section 230. It is unlikely that Common Carrier considerations were on the "desired outcomes" list.
     2. Example: Concerns have been expressed that MV is bad for editor recruitment
     3. Example: MV leads to viewing images without a thumbnail preview. This could impact the Gender Gap negatively, affect the consideration of WP as work-safe, child-friendly and respectful.

• 1. Do we think RfCs to be the best means of achieving a decision or consensus? -- 2. They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?

1. Yes, if they are appropriately advertised. The distinction between an open-ended RFC and a !vote based one are covered by one of our German contributors.
2. This depends on the nature of the project. For software, sharing early mock-ups, prototypes and alpha software with the community is ideal. It is established wisdom in software development that the earlier a bug is caught the less the cost of fixing it, by orders of magnitude. (Where ignoring a bug alienates your customers this is probably worse.)

• 1. Does every decision need to be community vetted? 2. If not, where is the threshold?
  1. "Have to" no. Would it benefit from it? Maybe. A chance conversation saved a (fairly small) company I worked for £20,000 pa on its office stationery bill.
  2. I recommend you read Ricardo Semler's book Maverick to see how far this can be taken. (Or failing that the Wikipedia article en:Ricardo Semler.)
  • This is a red-herring though. The crux is that we need more respect of the community position, in order to have a healthy relationship. Secondarily the work of the Foundation and the Community will benefit massively from the outcomes of collegial working.

• 1. Does every decision need to be WMF vetted? 2. If not, where is the threshold?
  1. No. There are legal issues if WMF gets involved in content, for example.
  2. Setting aside content, it depends, there is no reason that the WMF, if it has the resource cannot give assistance in graphical design, template writing, Lua coding, governance, and training. Indeed I have specifically proposed that professional training for senior functionaries (broadly construed) in dispute resolution, ethics, and a few other topics would be valuable. I certainly see scope for some of this training to be made available to staff.

Communications

• 1. Where should we communicate to reach enough critical people so we don’t get into a place where we waste work? 2. Are we using the right communication channels to reach contributors? 3. Readers? 4. How do we prevent ourselves from jumping in at the very last moment (if we missed getting engaged on-time) when the works is ready and public?
  1. See above - discuss, research, discuss, mock up, discuss, prototype, discuss, alpha....
  2. It's hard to say - this would be a good research project in it's own right. Personally I miss most of these discussions, but generally the community has said the right thing, so I don't worry.
  3. Again research is needed but we can easily have a short reader survey filled in by thousands of readers. There are issues around user perception vs actuality, so it depends on what we want to learn.
  4. Who is getting engaged at the last minute? You and me, certainly, but the community and staff have been discussing this for a long time. Again, discuss, research, discuss, mock up, discuss, prototype, discuss, alpha....

So why weren't issues hashed out earlier? The problem I see here is that everyone seems to be OK until it's in production. Then people who did not care before wake up once a feature is live. This is not efficient and is conflict prone as well. So I do think we have a communication and agreement problem. There need to be checkpoints along the way -- like in your research/mocup/prototype. But they cannot be simply discussions, they need to criteria for moving to the next stage. -- LilaTretikov (talk) 15:56, 19 August 2014 (UTC)

• Who should be a part of the communication distribution? Should it be exclusively the Community, the WMF, or a combination of the two? If a combination of the two, how should responsibilities be broken down and shared?

• Why are personal insults a part of our methods of communication? How do we stop that?
  1. I think insults are likely when one party is stone-walling. So we shouldn't do that! But the question of "civility" is broad-reaching, and multi-facted. The two key things to remember are that written communication conveys much less "tone" than spoken communication, and secondly that we have a wide variety of friends here, form all groups and nations. Expressions and words have different meaning and intensity for different people.

No one should ignore or stonewall. But people should own and be responsible for their own behavior. Attacks and name-calling is not OK under any circumstances IMO. -- LilaTretikov (talk) 15:56, 19 August 2014 (UTC)

• 1 How do we stop pointing fingers? 2. How do we align ownership, execution and authority without blocking all changes?
  1. If actions are taken with consensus finer pointing becomes untenable.
  2. There is no reason to think that the community is change adverse (at least not from the high-profile cases) it is simply against bad ideas (image filter) and releasing software that is not ready. It is true that in the case of Pending Changes what was seen as an attempt to force it through effectively scuppered what might have been acceptable to the community if it had been handled better. For the English Wikipedaa as far as most software changes are concerned there is a flurry of activity on Village Pump (Technical) as bugs are reported, or a new feature is explained, and little or no dissent. "Thanks" for example met with broad community approval, I believe. Had the community hated it, it could have simply been disabled - this is not a bad model for low-effort low-cost software improvements. It works too for big projects, if we are prepared to throw them away (or fix them) if they fail.

We definitely need to have a better methodology for feedback and that's a priority. -- LilaTretikov (talk) 15:56, 19 August 2014 (UTC)

Planning for the future

• Is it possible for us as a community as a whole (and as many smaller communities) to think beyond what we do today: editing long-form Wikipedia articles? If so, how would we do this? I'm not saying this because I don't think we should work on making long-form editing easier, I want us to recognise that there are more types of content and ways to contribute and consume it today.
  1. Yes - and we do, key activities include
     • Outreach
     • Media (mainly images, but also sound, video and other content)
     • Translation
     • Data
     • Metadata in content
     • Research

• What external threats should we be planning for?
  1. Commercial competitors
  2. Legal extinction (class action lawsuit for $1bn, for example)
  3. Loss of net-neutrality
  4. Search engines downgrading our content
  5. Attacks on volunteers (legal or otherwise)
  6. Government censorship
  7. Regulatory burden
  8. Technical attacks from capable opponents

What we should not be worrying about at present is "Right to be forgotten" - this does not impact current projects. Certainly there are long term issues, relating to what I shall loosely call Big Data, but nothing for the next five years or so.

Practices in Technology

• WMF is a technology org. Do you disagree with that mandate?
  • Partially. Technology delivers information. But information is delivered for the benefit of, and mainly to, people. Therefore we coudl justas easily say
    • WMF is a knowledge org.
    • WMF is a people org.

• Our current change-management practices around Site Operations/DevOps are dangerously out of date (especially how we change production code). How should we fix that?
  • I'm not sure of the current practices, so I offer these suggestions form a basis of general good practice
    • Build development and user acceptance mirrors - and use them!
    • Always have a revert plan
    • Write orthogonal code
    • Pilot!
    • Plan phased releases of larger components

• 1. What did you like that WMF development has done? 2. What parts of the process worked? 3.Which parts of the process did not work specifically, and why?
  1. The thank feature, revert monitoring, ping.
  2. & 3. Which process? I don't watch the development process, and I'm not sure if there are tools that would allow me to do so, short of looking at actual code.

Readers & Editors

• 1. To what degree should we care about the readers? 2 Reach?
  1. 100%, though not all readers are human, not all readers are current readers and the community is made up of voracious readers
  2. Reach is important, and we have known for at least three years largely mobile driven. I advocated at Haifa that as well as Wikipedia Zero, we should encourage mobile phone manufacturers to include a snapshot of Wikipedia stored on the phone. I'm not sure if any progress was made with this, but certainly there's scope for a static or cached version to sit on tablets and other devices (book readers being the obvious first target).

• 1. Should we split the site into reading and editing experience? 2. New users and contemporary?

1. Having a shared experience is believed to draw in new editors. We have, though, the "edit" tab, so there is some distinction. The Visual Editor attempts to blur this, but I haven't seen any research to suggest that it is a net benefit (or otherwise).
2. Isn't this want happens with mobile? I'm not sure it's desirable, but maybe it is unavoidable.

• Do we believe that the expectations of both readers and contributors have changed over the last 10 years? Our UX studies show this to be the case: the users who come to Wikipedia today get confused because they have been trained to expect different UX patterns. How should we handle this?
  1. Yes, but I think people still largely expect to "read" an encyclopaedia article.
  2. Do the UX studies compare what people expected 10 years ago with today? If not they can't "show this to be the case". Link to UX studies?
  3. Depends on the cognitive dissonance MW interface is creating.

• Do you think our mission is just about editors? Without my making any judgment here, would you write if no one were reading?
  1. It depends what you mean by "editors". If you mean those who "collect and develop educational content under a free license" and who "disseminate it globally" then yes indeed.
  2. I would still write, for my own consumption, but the subjects might be different. I think also it's worth realising that even if no one read another page on Wikipedia our content is syndicated through Google, Facebook and thousands of other sites, provides meta-data to an unknown number of projects, and a rich resource of images, translations, definitions and much besides.

Features

• 1. Do you think old features need to be redone or removed? 2. What would you do to clean up old cruft and features? 3. How would you tell the few hundred editors that would get really upset?
  1. Depends on the feature. Give some examples.
  2. Cruft is an emotive word. Generally though it can be rationalised into a more universal interface element. Old features - such as categories, would benefit from some work (search within category, category combiner, tagging, per-category sort order), but have not stood still (PAGEESINCATEGORY, category intersection, DEFAULTSORT, change in sort order, tree displayer). Certainly if many new sub-features have been added there is probably scope for a code-refactor.

• What are the top 3 product features that you would like to see rolled out over the next two years for readers, assuming we could meet your quality standards?
  1. Faster page load times (note this requires removing cruft from the huge java script and css downloads from the "bits" server, cruft in the html, cruft in the site js and css, as well as potentially refactoring templates and lua code, and ensuring we serve compressed HTML to clients that can read it). As a side benefit this could free up much bandwidth.
  2. Natural language query.
  3. Closer integration of the projects, e.g Wiktionary mouseovers on Wikipedia

• What are the top 3 product features that you would like to see rolled out over the next two years for editors, assuming we could meet your quality standards?
  1. Reasonably efficient interwiki transclusion
  2. Improved Commons search integrated with tagging, ontology, structured data, Wikidata, and categories
  3. Dynamic category trees based on declared categories (would also require a distinction between sub-categories and member categories, or a clean category tree). Use cases 1. male and female X categories without ghettoization, 2. stub categories based on the main category tree.

I hope you find these responses useful. Rich Farmbrough 12:55 19 August 2014 (GMT).

Brief answer from Colton Cosmic to the first grouping of Lila's Questions

Hi, you asked, on Decision-making,

"1) How do we make decisions that move us forward, rather than backwards, when we cannot all agree? When the easiest choice is the status quo?" That is leadership isn't it? So, your problem. I would say that leadership implies going to a destination, so status quo is not leadership.

"2) How can we reach representative decisions for all wikis? Good practices in product tell us that we should test for desired outcomes: before and after to tell us if we succeed – that provide specifics on what needs to be fixed. Do you agree with that?" You should develop a bot that randomly polls the complete list of editors if you want a scientifically representative gauge of what all the editors and administrators want. Legotkm has a bot that could be adapted to do that. It gets harder to figure out what the *readers* want.

"3) Do we think RfCs to be the best means of achieving a decision or consensus? -- They are great for putting forward ideas and pointing things out (IMHO) – what tools do we need for other phases of projects?" No. RFCs as they exist now are only good for determining the viewpoints of the small set of editors that participate in RFCs. They are not scientific as far as gauging the community's viewpoint, and I think their closures are often warped and manipulated by involved parties.

Colton Cosmic (talk) 13:51, 19 August 2014 (UTC)

Colton Cosmic thank you for your feedback -- especially #2. I have been recommending internally to post messages to talk pages or email a % of registered users to participate in tests. I am going to move that comment to the idea page. -- LilaTretikov (talk) 14:05, 19 August 2014 (UTC)

Old templates

"We have some that are very very old, including inconsistent old templates, but some people just love them. It's expensive to maintain."

The community maintains the templates. And the "cost of maintenance" is often a myth, I have demonstrated (but not proved) this with redirects, and certain families of templates.

If you can give me examples of these "inconsistent old templates" I would probably be able to evaluate wheterh there is any sense to this surprising statement.

All the best, Rich Farmbrough 13:12 19 August 2014 (GMT).

That's the worst spell of weather we've had in a long time. Colton Cosmic (talk) 13:33, 19 August 2014 (UTC)

It is a good point that this may be best done by people who maintain them now. We just need to organize on what we want to do here. Here are the issues I see from my very high-level perspective:

There are a lot of them, some duplicate/overlap in function

As we move towards more modern interactions (like with VE), accommodating for editing templates means adapting to each one individually as they are non-standard (hence the cost)

They all look very different, in style, design and layout, which give the site a strange and sometimes unpredictable look (when we want it to look very clean and encyclopedic)

But you are right.. we need to take inventory, decide what we want, then try to problem solve. I called those out because they are presenting problems with modernizing editing. -- LilaTretikov (talk) 14:02, 19 August 2014 (UTC)

"As we move towards more modern interactions (like with VE)" - that proposition seemingly tacitly includes value judgments that might not be shared by those who might be in a good enough position to judge where "we" should "move" -- which are, naturally, for each individual project and community context, foremost the respective most experienced contributors there. Especially, while general compatibility with VE would be nice-to-have, it is for many applications (with largely shared exceptions, e.g. where table formatting is concerned) very far from being a key concern regarding where "we" should move. However, it is hard to understand what we are talking about in this context at all, until we have examples for which "templates" might be meant here. You probably are using the term metaphorically instead of technically here, aren't you? Now, could you please provide examples? Ca$e (talk) 14:49, 19 August 2014 (UTC)

Well since Bing engineers just contacted OTRS about how the hell they are supposed to get consistent geo data out of our 'templates', I think that that is where we should look in this regard. They might not be expensive to an experienced article editor, but they are incredible expensive to ppl outside of that group. Also for instance for something like Parsoid. The fact that our entire wikicode language is an adhoc approach is causing us (Developers/WMF/community/movement) pain everywhere for years already. For a community so intent on sharing knowledge, when it comes to actually being able to disseminate that knowledge, we have built a behemoth of inability on that front. As GLAM collaborators have often told me: getting images into MediaWiki/WM is a pain, but getting information out of MW/WM is even worse, and many of the causes for that are that we have found brilliant ways to get around limitations, that are only sustainable to ourselves and not for anyone else. Even as an experienced template editor, I can attest, that I have chosen not to cleanup templates on en.wp simply because I didn't want to understand all the things they were doing currently. Rather I let it rot and pollute the wiki longer, in favor of not breaking it in the short term. TheDJ (talk) 15:51, 19 August 2014 (UTC)

Alanscottwalker on Software RfC's

Do we think RfCs to be the best means of achieving a decision or consensus?

Not on project-wide software placement. Request for Comment (note the name), in ideal, are for reaching consensus, and consensus is not majority decision, it is taking all valid concerns into account and coming up with something everyone can live with under agreed valid standards. This is sometimes called in shorthand "votes are evil." Voters in RfC's are self selected - not representatives, this may work for discreet content issues (where metrics for decision making are clear) but not for turning off project-wide software. The voters may, in majority, sometimes agree on result but, as here, for a multiplicity of reasons, many of those reasons invalid -- thus there is no real consensus. If there is legal objection to software, that's not something to have a vote on by mostly anonymous, inexpert, and legally unaccountable people. If there is emotional objection to software, that's not something others can vote on, because we can't vote on others' emotions. If there is technical objection, those are often addressed/changing during the vote, meaning the votes are on different things and not on the same thing, which is what you would need, at the very least, for forming any consensus. If there are purpose objections to software, and some see it as a valid purpose and others do not, that is not a consensus. If there are means objections to software (that is, the means reach the purpose), and some say it does and some say it does not, that is not a consensus. If there are entirely different metrics/standards being applied by different sides, that is not a consensus. (There maybe some use somewhere for small questions/issues RfC's in the software development process, but not like we have generally seen) -- Alanscottwalker (talk) 14:47, 19 August 2014 (UTC)

Working Together

Thank you everyone for the insights and arguments you’ve shared with me and each other over the past week -- I am continually using them to inform my thinking. Your passion is undeniable, and I want you to know that my passion for this movement is too. I’ve read what you’ve sent my way these past few days. In return I want to share with you my thoughts on where I think we are, and where we need to be.

The Board brought me into this project in order to lead a transformation at the WMF. I accepted the challenge because I believe you’ve created an incredible project that has changed the way the world experiences knowledge. However, I also believe both your contributions and our movement’s mission are at risk of being lost as a result of changing technologies, much as Encyclopedia Britannica was in the 90s.

This means we must make some changes. And we must make them together. I spoke about this at a high level at Wikimania, but in action this means we must:

• Improve our process for software design, user and community feedback, and operations;
• Learn how to work together through disagreements and make decisions that have global impact with objectivity;
• Improve our product consistency globally and think of ourselves as the world’s source of free knowledge;
• Increase the responsiveness and speed at which we develop and deliver product; and the rate of innovating; and
• Do this with mutual understanding and respect.

This weekend I posted a set of questions here, on my talk page. They represent the issues I am grappling with, and the things I wish to understand in order to assess the specifics of changes we need to make. I appreciate the responses you’ve shared so far, and look forward to receiving more. I would like to quote Martijn Hoekstra, who stated on this page that it is “the burden of the person with the initiative to initiate the dialog. In case of WMF software projects, that's the WMF.” We are going to do that.

I want you to know that I hear you, as different as you all are. Everyone has legitimate concerns about the current situation: not just the recent issue involving the WMF and the de.wp community, but the ways in which we work together overall. I actively engaged with you to solve this. We need to transform our conversations into an ongoing and improving process with common goals.

While we are all part of a larger community, we have different tasks. Of our movement entities, the Wikimedia Foundation is in the unique position to lead the continued development of the technology making these projects possible. This means we understand ourselves as a technology organization. This also means we are a global organization responsible for technology powering 800+ projects.

Yet many of you need to be able to influence and affect the direction our projects take -- this means being heard and taken into account even on things you don’t directly work on. We need to find a way for all of the contributors from all of those 800+ projects to participate. You need opportunities to review the development of product earlier and during critical junctures. Our approaches need to evolve and mature. We need to find a better way to collaborate.

The WMF is a part of our community -- the part that is responsible for developing and maintaining software and servers. We all want to want to participate in deciding which features get implemented and how, but the current approach of voting on them post-rollout is disruptive and inefficient. We need to change our processes so that they are iterative, incremental, and inclusive of feedback throughout. We understand that our recent decision to restrict edits to site-wide JavaScript on German Wikipedia was a surprising move that upset a significant number of people - we’re sorry for that. At the same time, it gave visibility to an important issue: we need a better mechanism for managing changes that impact all users -- the way the MediaWiki: namespace works right now is not sustainable. Let's use this opportunity to improve.

At the WMF, we’re preparing to unprotect the disputed page on German Wikipedia, but we need to do so within a framework that allows us to come to reasonable resolutions -- giving everyone a voice and a say in the process, but also understanding WMF’s leadership role in technology. We will post thoughts on how this can be accomplished (and in what timeframe), in partnership, in coming days, starting with a https://meta.wikimedia.org/wiki/Community_Engagement_(Product)/Process_ideas#Notifications_for_beta_features.

As part of this process, we have heard feedback that WMF employees should have distinct accounts for their WMF-related actions as opposed to their personal actions on the projects. We accept that feedback and will put in place such a system within the next month.

In summary:

• You have my commitment that we will work towards a constructive resolution of this current and any future disputes together and in good faith.
• We intend to undertake a review of our present processes immediately and propose a new approach that allows for feedback at more critical and relevant junctures in the next 90 days. This will be a transparent process that includes your voices.
• We will establish improved centralized communications for all wiki software changes.
• All future updates and current developments will be based on this new process.
• For the purpose of additional clarity regarding roles and responsibilities, we will put in place a clear distinction between work and personal accounts for all WMF employees by September 15.

I hope we can agree to exercise restraint in this transformative time so we can work together in good faith and in concert. As Magnus Manske said in his recent blog, “the house that is Wikipedia cannot stand without a foundation, and a foundation without a house on top is but a dirty pond.”

Thank you, -- LilaTretikov (talk) 16:22, 19 August 2014 (UTC)