Talk:Community Liaisons/Process ideas

Add topic
From Meta, a Wikimedia project coordination wiki
(Redirected from Talk:Community Engagement (Product)/Process ideas)
Latest comment: 7 years ago by Rogol Domedonfors in topic Historical

Thank you for your involvement![edit]

Thought it might be helpful to start the Talk Page. Thank you to everyone posting your ideas. -Rdicerb (WMF) (talk) 05:48, 20 August 2014 (UTC)Reply

Hi Rdicerb (WMF). I would like to echo what Sicherlich said below. This page is quite confusing and I think that is mirrored in the number of edits its getting. You should realize that a large majority of users commenting on this might be from de.wp and might not have English as their native language. It is not surprising users seem more comfortable leaving messages anywhere else than here. It seems Lila's user page and RfC are still the favorite. I'll try and offer some constructive criticism, it is only intended to help.
The lead section offers little information what this page is for, and instead spends half its length laying down guidelines. You can assume if people are reading it here, they have more than the average grasp of editing behavior and talk page use. Also, coherency isn't something that you can just ask from a crowd giving its opinion - the formulation of a coherent feedback page should be separate from soliciting the said feedback.
My biggest issue is, there is absolutely no way to see who said what, save for searching through history for an hour. There are no signatures, so I have no idea if that was pre-filled by a staff member or if it's someone's opinion. It really gets confusing the more you try and follow it.
You are by far the largest contributor to this page. The lead section and the page name suggests that this is for community brainstorming but so far you, James and a few other staff are prob. responsible for almost half the edits on this page.
I have a slight issue if the staff members get an overwhelming weight here to voice what they think - I believe we are in this predicament because of feedback pages like this? I have no way of knowing if the ideas here were pre-selected by the staff and they proceeded to add what they think is an advantage or disadvantage - a process like this becomes more of a fishbowl exercise and doesn't require outside intervention if you don't let the majority in to the process of formulation.
This page definitely needs better organization and perhaps branching off in to multiple pages? It's too long at the moment and its hard to follow anything. The section at the absolute bottom "WMF- getting to know Wikipedia-community" is the only one that makes sense or seems straight-forward, the rest is too abstract and poorly organized to follow.
Perhaps, you might consider a survey? It seems if these are pre-filled that you want to ask people some question instead of discussing them since everything is anonymized to a certain extent.
Anyway, That's it. I hope this feedback helps in some way. You can fix most of these things, and you should probably consider a survey instead if this is the format you want to follow. Theo10011 (talk) 20:10, 21 August 2014 (UTC)Reply
+ 1 -- I came here to offer my two cents of brain. But the page does not encourage me to do so. Discussion takes the guise of non-signed advantages/disadvantages -- very non-wiki.---<(kmk)>- (talk) 20:55, 22 August 2014 (UTC)Reply

POV in "Gaps in the process"[edit]

Despite being a member of the community and writing this message in that role, I'm not going to edit this directly because I fear some people would accuse me of acting in WMF's interests.

Reading through the many additions to the "Gaps in the process" section, I see a few that appear to be POV, assuming certain conclusions that are IMO not warranted:

  • "A person has a voice of his own. Allow the person to voice, and respect quiet as agreement with what others decide." – Silence is not always consent. It can as easily be that they didn't know about the discussion, or that they didn't feel empowered to comment. While I agree that better studies are needed to actually determine what readers want, just assuming that vocal people (on either side) are representing their views seems flawed.
  • "There is a lack of understanding that votes are a proven deescalation measure. so do make votes, and respect them. if the vote does not go your way, the question or the proposal was not good enough." – Blatant pro-voting POV. Voting isn't the answer to everything; sometimes experts need to be trusted to make the decisions. Note that "experts" is not the same as "WMF staff", and that most decisions could use input from many people who are each experts in individual aspects of the question.
  • "there is statistics and further strong evidence that the processes around wikipedia are too complex and heavyweight. despite this, every "it is not going my way" is followed by introducing a new and more heavyweight process." – Blatant anti-superprotection POV, and seems likely to be untrue.

Anomie (talk) 14:06, 20 August 2014 (UTC)Reply

Thank you for bringing this up. This is a strongly felt topic and I doubt we can request everyone stay strictly NPOV at all times, but I'll nudge people in that direction. -Rdicerb (WMF) (talk) 17:48, 20 August 2014 (UTC)Reply
I see we have a new one this morning in "Current state": the bullet beginning "When the WMF prematurely releases software" has some valid points, but they're lost in an anti-WMF screed. Anomie (talk) 13:49, 21 August 2014 (UTC)Reply
It's actually not an "anti-WMF screed". I do think that a description of the current reaction of the WMF when their software is disabled needs to be mentioned, and it unfortunately does range from threats to blatantly lying. The focus here seems to be all aimed at how to come to an agreement, but doesn't deal with the unfortunate reality that sometimes agreement will not be found. We need to focus on what kind of behaviour on the part of the WMF is acceptable when a community adamantly refuses to tolerate the implementation of WMF provided software, and the current strategies certainly aren't acceptable, nor is removing the only line of defense the communities had by implementing "superprotect".Kww (talk) 06:12, 22 August 2014 (UTC)Reply
I also see someone added weasel phrasing in changing "don't represent readers directly" to "can be perceived as not representing readers directly" in this edit. Anomie (talk) 13:53, 21 August 2014 (UTC)Reply
If you prefer I can remove the bullet point altogether; the reader doesn't exist anyway. --Nemo 06:24, 22 August 2014 (UTC)Reply

Moved from Process List[edit]

I'm moving these two out of the process list, as I think they have less to do with the process, and more with design principles in general. Documenting design principles seems like a good idea to me, but they feel out of place to me in the list above...... might have to be moved into a section below as a separate section ? TheDJ (talk) 12:17, 20 August 2014 (UTC)Reply

  • There is a lack of "one click more" "is a billion clicks in total for all the people" understanding especially at WMF designers and developers, as well as "1 sec more" "is a billion of seconds for all".
  • There is a lack of understanding that there is two user groups, the editors and the readers. the editors think the readers can be ignored as they come anyway reading, no statistics support that we do not have enough of readers, or loosing readers. contrary, we have statistics supporting we do have difficulties attracting and keeping contributors.
Hey there, thanks for keeping it focused. Alternatively, we can also put off-topic comments and suggestions in the Out of Scope section. -Rdicerb (WMF) (talk) 21:53, 20 August 2014 (UTC)Reply

Moved comment from gadget improvement section[edit]

Moved from the main page; please don't try to turn it into a discussion piece. Jdforrester (WMF) (talk) 17:48, 20 August 2014 (UTC)Reply

I'm a big fan of this. Better documentation of our JS components, a default ui toolkit (OOJS/mediawiki.ui), better search interface for gadget users, integrated translations, templating, integrated reporting of errors that are traced to the gadget tool, review etc etc. Interestingly enough this was being worked on in the past by Roan and Timo, but we needed them on VE. Then again, volunteers should be able to pick this up (why are they not ?? do we need to invest into those areas first ?) Currently many of the smaller requirements for this (templating, ui toolkit etc) are being driven by VE teams, perhaps we are getting closer and it will already be easier volunteer devs to achieve this ? TheDJ (talk) 12:44, 20 August 2014 (UTC)Reply

Communication channels[edit]

Hello, I am not sure whether this is the right place, or what you already have done so far in SF. In WMNL, I once made a list of all communication channels that came into existence earlier. Then we selected some (website, mailing list, direct mail / newsletter) and explained on a General Assembly what kind of information we will put where. In the meanwhile we have also deleted some channels.

Is there such a list for the WMF, do Wikimedians find it, and does it say what kind of information a peticular channel delivers to what kind of readers? This might help the WMF to make sure that no topic and no target group is overlooked. Also, maybe the channel for technical stuff is realistically only suitable for tech folk, while for usual editors that information is unintelligable.

On the other hand, I think that also Wikimedians have an obligation to actively look out for information they deem to be important (if the WMF provides a suitable channel). I remember a WMNL member who once bitterly complained about the communications of WMNL because it did not inform about something. I then pointed out that that information appeared in our newsletter. His answer: "Ah, there is so much." (Actually our monthly newsletter is just 1-2 screen pages long.)

I sometimes see a certain tendency in an organisation that it wants to open many channels to please different kind of people. For clarity and the ease of maintenance, I came to believe that very few channels are a much better solution.

Kind regards Ziko (talk) 18:40, 20 August 2014 (UTC)Reply

Hi Ziko, this would be the right place to bring up this sort of comment, as it involves how users are notified. You are welcome to post any ideas as such in the main page. Thank you for posting that list - there are several lists of wikis and Village Pumps and message lists etc on Meta as well. While I agree with you that information is posted throughout wikis, from my perspective it seems almost like information overload that isn't always very easy to find (at least, not for someone who is trying to learn!). I wonder how everyone could find ways of centralizing some information while managing the burdens that would come with it (annoyance, information overload, turning off the notifications and "not getting the memo", etc). -Rdicerb (WMF) (talk) 00:20, 21 August 2014 (UTC)Reply
@Ziko: There was a roundtable discussion about this at Wikimania, with these notes on the various communication channels and test mechanisms. I'll copy a cleaned-up version of that list to here:
and everyone came to much the same conclusion (that fewer channels would be better, in the long-term).
Also comments about: Push Notices tend to suffer from en:banner blindness.
How Echo Notifications have been (purposefully) under-used for the last year to avoid banner-blindness and en:information pollution, and especially because SUL finalization is blocking Global Notifications (we don't want to each get dozens of identical notifications telling us about an upcoming software change, from each wiki), but we should eventually be using that system more, plus adding more opt-in interest-groups, and possibly newsletter-signups. There were suggestions about creating new low-quantity echo feeds, and slowly suggesting it to more and more active editors (plus widely-announcing it to get additional opt-in signups).
Also comments about how few (relatively speaking) of us are subscribed to the community newspapers (eg. 850 at Enwiki subscribed to The Signpost, though many just check it weekly without a subscription, or via noticing someone else's in their watchlist).
Lastly, there are further lists at mw:Movement broadcasting, and en:Wikipedia:News. (Sorry for the poorly-organized deluge. It's been a long day. :) Quiddity (WMF) (talk) 04:41, 21 August 2014 (UTC)Reply
Quiddity (WMF): I see that Tech/News isn't listed among the channels above, or in the Wikimania etherpad; it may be an oversight, or simply a sad symptom showing how little it's known, despite our efforts to advertise it :( It was created precisely to avoid tech information overload and we notably use it to let people know about early designs so they can provide feedback. guillom 16:05, 21 August 2014 (UTC)Reply
guillom: The list was just a brainstorm during the roundtable (which was leaning towards a focus on design collaboration/communication) (and partially fleshed-out by me last night), I didn't mean to imply that it's complete. However, I would also suggest that Tech/News could be considered a part of the "newsletters" group, and also the "mailing lists" group, as it gets delivered via both those systems ;) (But yes, it could benefit from more visibility, contributors, subscribers, etc). I know there are some good ideas around revamping the newsletter-system, but as with most good ideas, they're awaiting time/resources to bring them past the idea-phase (and to do it "right", rather than fast/simple). Hopefully soon. Quiddity (WMF) (talk) 18:14, 21 August 2014 (UTC)Reply

Hello, thanks for the information! I am not sure what to do with the idea for the process, except creating just another communications channel :-) I'd like to stress out that a channel must be really suitable for the target group. / In the old days, a movement or association had simply one magazine or newspaper, and every month you got the information someone deemed to be necessary for you. Not perfect, and it created some tension about who would get more lines in it, but maybe more efficient than what we in the movement are doing now... Ziko (talk) 16:56, 22 August 2014 (UTC)Reply

A note about translating[edit]

Hi there; we're going to clean up the "Advantages, Disadvantages" to make it easier to edit. After that Community Engagement will be marking the page for translation. -Rdicerb (WMF) (talk) 18:55, 20 August 2014 (UTC)Reply

Translating this simulation of community involvement is nice, however, as explained e.g. here, do not expect that many people will regard this as an acceptable platform for discussion. Ca$e (talk) 19:19, 20 August 2014 (UTC)Reply
Follow up to my comment - Marking the page for translation while people are brainstorming here is likely to become burdensome for translators. If you translate pages and disagree with this, please feel free to let me know, but I think at this point we won't be able to mark for translation on this page while content is being freely contributed. Ca$e, I do hope that people are interested in this and find it acceptable to post their ideas! -Rdicerb (WMF) (talk) 00:02, 21 August 2014 (UTC)Reply
Given the new vision of WMF for its "future" and "role", there is nothing to discuss anymore. WMF will do what it will on the way towards a top5-facebook/twitter-alike, will ignore former basic principles ("the 4th principle applies to some degree"), no matter what parts of the community will fork off, given that they are nothing but "the loud and the few" and (Jan-Bart:) "we cannot enact the wishes of a few hundred". Erik: WMF "needs to have a say in the _outcome_, even if/when there are RFCs/votes asking us to disable a feature" ([1]), "WMF needs to be able to make an ultimate Determination" ([2]). We will not take part in this simulation of community involvement. Jan-Bart: "if you decide to take a wiki-break, that might be the way things have to be. Even so, you have to let the Foundation do its work" - no we definitely do not. We invite you to discuss this brave new and "innovative" vision with yourself. Ca$e (talk) 07:43, 21 August 2014 (UTC)Reply
Ca$e, I think this is mainly an internal WMF brainstorming, conducted in the open (which is something worth praising, unless there is some parallel ongoing discussion behind close doors as well). If you consider that thousands of Wikimedia editors are discussing the matter in their "own" venues, it makes sense that the couple dozens or hundreds WMF employees interested in the topic also have a venue they feel comfortable with to elaborate some shared thinking. --Nemo 06:27, 22 August 2014 (UTC)Reply
And that would be alright, would they not at the same time simulate that this was about "community engagement". Even the page header says so. It is, as i laid out, plain ridiculous to use such terms in the context of the position of the WMF board. We will just not buy it anymore. Ca$e (talk) 08:02, 22 August 2014 (UTC)Reply


My edits to the page were in my staff role, and in the wrong account. Ironholds (talk) 22:00, 20 August 2014 (UTC)Reply

Just change the sig to reflect that, using three tildas, in front of the date stamp, to replace Ironholds with Ironholds (WMF) or whatever your staff sig is. Or just do an edit/replace. Apteva (talk) 14:21, 7 September 2014 (UTC)Reply

Title Case Is Disturbing[edit]

We use sentence case. MediaWiki is not UseModWiki. --Nemo 05:21, 21 August 2014 (UTC)Reply

Change boldly? Steven Walling (WMF) • talk 16:12, 21 August 2014 (UTC)Reply
ChangeBoldly —TheDJ (Not WMF) (talkcontribs) 18:07, 21 August 2014 (UTC)Reply
This one is on me, Nemo. I didn't realize and I'm still learning the culture of each Wiki. As others said, please change boldly! -Rdicerb (WMF) (talk) 22:45, 21 August 2014 (UTC)Reply
On the other hand, sentence case for page names and subject headings is sort of a bizarre left over from when Wikipedia was created by 5th graders who did not know any better. It was done so that when you create a link to a page in a sentence you could do it without using odd capital letters in the sentence, but the solution is to redirect from the lower case to the correct page so that it does not create a redlink, not create a goofy standard. Since section headings can never be linked to directly that choice was just spill over so that you could make an article that looked consistent - consistently wrong, but still consistent. All caps is "disturbing", or at least annoying and hard to read, but anyone who thinks that "title case is disturbing" had better get over that feeling, as it is standard practice in all of the world except on Wikipedia. Apteva (talk) 14:37, 7 September 2014 (UTC)Reply
"All of the world" must not include the BBC or Reuters. Whether to use sentence case or title case is a style decision, and different publications make different choices. If you look at this sample of English-language newspapers, you'll see that the overwhelming majority choose sentence case for headlines. It appears that w:The Times They Are a-Changin', at least with respect to whether sentence case is fashionable. WhatamIdoing (talk) 05:00, 8 September 2014 (UTC)Reply
Wouldn't it be funny if they only reason they did it was from editing Wikipedia when they were younger. Apteva (talk) 16:22, 8 September 2014 (UTC)Reply

Don't get it ...[edit]

Maybe my english is to weak, maybe I dont know enough about software development but I dont get that page. a lot of headlines, some questions, some "Prioritization process in a nutshel" some answers ...
Funny enough: while on other heavily discussed pages there are only few WMF people active here it seems to be them to be active but hardly Wikipedians!?! Maybe that shows the difference in the working culture? WMF does not get the Wikipedians while the Wikipedians does not get WMF? ...
I have no idea how to get to know WMF work (no I dont have time or desire to visit you :) ) i think WMF staff should learn about Wikipedia "inside" as this is what they are working for. Thats where the money the earn comes from. So I just droped my ideas. I have no clue if it fits to the topic. I dont know if it fits to the structure. If its totally out of place simply remove it. No hard feelings on my side :) ...Sicherlich Post 18:40, 21 August 2014 (UTC)Reply
Thank you so much Sicherlich. I will be more careful with simple English. Thank you also for your idea about getting to know communities. -Rdicerb (WMF) (talk) 00:14, 22 August 2014 (UTC)Reply

Sprecht ihr Deutsch?[edit]

Ein Problem was immer wieder auftaucht so auch hier. Sicherlich geht eine ganze Menge sehr wertvoller Input verloren, weil die Leute die gute Ideen und Lösungen haben kein Englisch sprechen. ... Barriere Nummer 1 - find meta und find diese Seite. Barriere Nummer 2 - sprich englisch. Selbst ich der seit vielen Jahren in der Wikipedia aktiv bin bin äußerst selten auf meta. Ich merke gar nicht wenn hier etwas passiert. Mein Englisch geht so um hier zu folgen. Aber viele sagen es ganz offen; Englisch ist nicht ihr Ding. Wenn ernsthaftes Interesse an Input von außerhalb der Englischsprachigen besteht, dann wird im ersten Schritt mehr übersetzt werden müssen. Ob sich das lohnt? K.A. Sicherlich nicht für alle Sprachen ...Sicherlich Post 18:48, 21 August 2014 (UTC) ja, und das ich hier Deutsch schreibe ist Absicht: Verdeutlichung am lebenden Beispiel quasi :) Reply

Sie dürfen in jede gewünschte Sprache schreiben. Whatamidoing (WMF) (talk) 19:55, 21 August 2014 (UTC)Reply
Ja; aber wenn ich kein Englisch verstehe weiß ich gar nicht was hier los ist. Was hier jemand will. Ob mich das Thema interessiert. Ob ich betroffen bin ...Sicherlich Post 08:48, 22 August 2014 (UTC)Reply
Perhaps we can make the introduction translatable ? I don't really have a solution other than that though. We need a common language at some point, perhaps we can translate the entire piece later on and have a second and third round ? —TheDJ (Not WMF) (talkcontribs) 11:44, 22 August 2014 (UTC)Reply
For discussions I dont see a good solution as well.
but if you ask for input; whom do you ask? People speaking english and hanging around on meta. But who is writing the Wikipedia? People speaking spanish, polish, german aso and hanging around on Wikipedia. - So if WMF is really interested in input of people who write Wikipedia 1st go where they are, and 2nd ask them in their language ...Sicherlich Post 11:56, 22 August 2014 (UTC)Reply
That's too easy. give specifics. go how ? central notices ? Translated in all languages ? And where do you want folks to post then here or on their own wiki ? —TheDJ (Not WMF) (talkcontribs) 12:45, 22 August 2014 (UTC)Reply
"Translated in all language" - in theory; practically it's going to be to much afford. So it's up to WMF how much they want to invest. How much it's worth for them to get feedback/input
"central notices" - IMO to much. But as already sad elsewhere; find out. On de its probably de:Vorlage:Beteiligen, maybe de:WP:K. probably in other languages elsewhere. If WMF does not know where to get in contact with the volunteers (or thinks that meta is a great place for that) this might be a small hint, that WMF has no clue whats going on at all. ...Sicherlich Post 14:28, 22 August 2014 (UTC)Reply
So you propose to hire contributors to work say ... 20-40 hours a week in a specific wiki to translate for the community, to spur the community into action and to keep eyes on what foundation does and how it delivers stuff etc.. ? Basically much what the community liaisons are already doing, but more of it and more focused on the community ?
In that case, I suggest the foundation gives a budget for 6 languages, a movement vote picks 6 languages, those language wiki's hold elections to select 3 candidates out of their community that they would find acceptable for that position. A 2nd round election would pick the final person, per language. This person is given a monthly grant by the foundation for the duration of his elected term. A term will last for a year. The next year will see new elections and the foundation can veto a reelection. —TheDJ (Not WMF) (talkcontribs) 17:40, 22 August 2014 (UTC)Reply
I dont propose anything. I just point out; If you want to talk with that would IMO the way.
If there is a need to hire people? dunno: there ist much more staff working for WMF then 8 years ago. Adding the chapters its a even greater number. While there are less Wikipedians and less edits .oO ... There are, at least on de there is, Community advocates. No idea what they are doing ... But both things are a different topic. ...
...Sicherlich Post 18:08, 22 August 2014 (UTC) anyways: honestly I dont think WMF is to eager to get too much involved with the community. So thinking how to achieve this is a waste of time Reply
Das sollte selbstverständlich sein, das die WMF einen gehörigen Teil ihrer Zeit in den Communities verbringt, insbesondere im Kern dieses "Unternehmens" hier, den WPs, um mit ihren Kunden, sprich uns, in Kontakt zu sein. Die werden bezahlt dafür, die haben aktiv den Kontakt und Rückmeldungen zu suchen. Sollte es eine Situation geben, in der die WMF etwas tut, das von den Communities abgelehnt wird, weil es nicht ausreichend diskutiert wurde, dann ist das ausschließlich die Schuld der WMF, sie hat dafür zu sorgen, dass alle Communities adäquat beteiligt werden. Die Communities müssen nicht ständig in iirgendwelchen komischen apokryphen Seiten und Listen suchen, die WMF hat zu liefern. Die WMF ist schließlich vor allem eine Serviceorganisation für die Communities, in keinster Weise ein Wert an sich.
It should be a no-brainer, that WMF staff spend a lot of time in the different communities, especially at the heart of the operation, the WPs, to connect with it's customers. They get paid by the communities, it's their duty to actively go to the communities for feedback. If a situation arises, where WMF did something, and the communities rejected it because it's not sufficiently discussed, it's the sole fault of the WMF, they have to make sure all communities are adequately involved. The communities don't have to look in some apocryphal pages for info, the WMF has to deliver it. The WMF is first and foremost a service organizsation for the communities, it's nothing on it's own.
--♫ Sänger, superputsch must go (talk) 07:40, 2 September 2014 (UTC)Reply
This does not seem very logical. Imagine that I see you on the street. If I ask you a question, and you refuse to reply, then you are saying that your choice not to talk to me is my fault. I'm inclined to say that your choice not to talk to me is your choice (and one that you should be free to make, so long as you are willing to live with the consequences). Whatamidoing (WMF) (talk) 16:44, 2 September 2014 (UTC)Reply
<quetsch> Die WMF ist eine Serviceorganisation für die Communities. Wenn die WMF etwas tun möchte, und die Communities nicht adäquat einbindet, z.B. durch aktives Fragen in der richtigen Sprache an der richtigen Stelle, mit anschließendem Warten auf Antworte hat sie bei Ausbleiben solcher, dieses als Ablehnung nehmen. Wenn die WMF also irgendwo in der deWP etwas auf Englisch fragt, von dem sie meint, es sei irgendwie wichtig, und dann keine Antworten bekommt, dann ist das i.d.T. die Schuld der WMF. Die haben sich gefälligst gut auszukennen in den diversen WPs, das sind schließlich deren Kunden, für die sie arbeiten und die sie bezahlen.
<squeeze> The WMF is a service organization for the communities. If the WMF wants to do something, and doesn't involve the communities in an appropriate manner, like active going there and asking questions in the right language at the right place, with waiting for answers, it should take the lack of answers as a no). If the WMF asks somewhere on deWP something in English, that it deem somehow important, and doesn't get any answers, it#s in deed the responsibility of the WMF. They have to know where and how to ask in the different WPs, as that are their customers, that they work for and that pay them.
--♫ Sänger - Talk - superputsch must go 14:48, 7 September 2014 (UTC)Reply
Do you really believe that, for every user-facing change, there should be a separate discussion at each of 800+ projects? Remember that there are 250+ languages involved here, and that we are talking about all the projects, not just Wikipedias. Or is it only for larger Wikipedias that you believe this should be done? Whatamidoing (WMF) (talk) 05:06, 8 September 2014 (UTC)Reply
Perhaps not really every language, for example Plattdüütsch, Alemannisch, Nordfriesisch, Boarisch, Deitsch, Nedersaksies, Pälzisch, Seeltersk and probably as well Lëtzebuergesch can all be made aware in German, you can use Dutch for Afrikaans, Nedersaksies, Frysk, Limburgs, Lëtzebuergesch, West-Vlaams, Zeeuws, you can probably use castilian as well for wikis in the other langauges of spain, if you ask politely and so on, so it's not really 250+ languages, but probably far less then 100. If you ask really politely in english, french, cantonese, russian, german and a few other languages, that suit that area quite well, this will probably do the trick as well, and someone will either translate it, or it will be understood first off. --♫ Sänger - Talk - superputsch must go 17:01, 8 September 2014 (UTC)Reply
So you believe that only 10 or 20 languages would be sufficient, plus requests for local translation (in both directions). Do you believe that 800+ separate conversations should be held? That is implied by "actively going there and asking questions in the right language at the right place". Whatamidoing (WMF) (talk) 23:02, 12 September 2014 (UTC)Reply
Imagine that I ask a member of the WMF staff a question, and they refuse to answer it. Is that my fault? [Hint: this is a trick question]. Deltahedron (talk) 16:53, 2 September 2014 (UTC)Reply

I observe another problem: The front page seems to appear very complex. A table of contents with about 25 sections is very deterrant for participation, and people very likely will add their ideas very unassertively. --Matthiasb (talk) 12:54, 22 August 2014 (UTC)Reply

Better this way ? —TheDJ (Not WMF) (talkcontribs) 13:33, 22 August 2014 (UTC)Reply
  • Piloten sind alle erforderlich, um zu verstehen und sprechen Englisch, so dass es zu einer gemeinsamen Sprache. Wir haben keine solche Anforderung. Ich würde Meta nicht behandeln, als Englisch, auch wenn die meisten der Beiträge sind in Englisch, aber behandeln sie, als ob es in Ihrer Sprache war, und fühlen sich frei, um in Ihrer Sprache, was immer es ist zu schreiben. Es kann in andere Sprachen, sowie Englisch übersetzt werden, aber es ist absolut nicht erforderlich, dass jede Seite auf Meta beginnen in englischer Sprache. (Blame Google for the poor German) Whatamidoing (WMF) is correct. You may write in any language on Meta. There are no exceptions, and no there is no requirement for a "common language" on WM. Thinking there is misses the whole point of Meta. Apteva (talk) 14:24, 31 August 2014 (UTC)Reply
    • Sicherlich is correct, though, that if you can't read the language that the discussion is written in (and perhaps don't know how to access machine translation), then it is very difficult to participate in that discussion. What do people think about putting a link to Google Translate at the top of the page? It's not perfect, but it's usually better than nothing. Whatamidoing (WMF) (talk) 23:16, 31 August 2014 (UTC)Reply
      • No. Google as a commercial product can not be treated as if it was a WM sister project. It only translates well into English anyway, and is relatively useless in translating into other languages - when it is used people complain that it creates nonsense translations in un-identifiable language instead of for example, German. There are other translators, and anyone who wants to translate can either find Google or the other translators without a link.
      • Sollte es ein Link an der Spitze von Google Translate? Nein. Google als kommerzielles Produkt kann nicht behandelt werden, als ob es ein WM Schwesterprojekt. Es übersetzt nur gut in Englisch sowieso, und ist bei der Übersetzung in andere Sprachen relativ nutzlos - wenn es verwendet wird, die Leute beschweren, dass es Unsinn Übersetzungen in un-identifizierbaren Sprache statt zum Beispiel Deutsch erstellt. Es gibt auch andere Übersetzer, und jeder, der übersetzen kann entweder Google oder finden die anderen Übersetzer ohne Link will.
      • Yesterday I noticed one project that had used Google translate to create an article in that language, and it was tagged to please improve the translation (and remove immediately? remove what?) Apteva (talk) 14:14, 7 September 2014 (UTC)Reply

Community workgroups[edit]

I like the idea of community workgroups with some modification. I'd really like to see individuals joining groups interested in a subject area - like editing, or multimedia, or search - and using that group to drive development. I would like those individuals to not be limited in joining the group if they would like to, and I would like it to be a requirement that you are a part of that group if you're working on software in that area. This requirement would apply to volunteers and staff alike. The group could come to decisions about merge rights on any repositories, and give feedback on proposed changes to the software. Basically, I want to see WikiProjects implemented for our software process.

I see this as the natural extension of my opinion that staff and volunteers are all in one community, and just have different duties and areas of interest. I don't want to see us expand the gap between community and staff, and I see this as a step in the right direction.

Realistically, this would need to be parallel to our existing (and evolving) software engineering process, but it could be made part of it relatively simply, I think. --MarkTraceur (WMF) (talk) 23:56, 21 August 2014 (UTC)Reply

OT comment: You know, this exactly summed why I am going away. 'You community folk need to organize yourself [thusly] to work with us, although we will still go ahead and do what we're doing without change.' Once upon a time things developed organically from the community. - Amgine/meta wikt wnews blog wmf-blog goog news 14:06, 22 August 2014 (UTC)Reply
@Amgine: I see this as much as a reorganization of our workflow as a reorganization of the community around our tech work. We don't really have a good system right now, and we need to reorganize no matter what, but this option allows us to meet somewhere in the middle (I think, anyway...) --MarkTraceur (WMF) (talk) 16:06, 22 August 2014 (UTC)Reply
@Amgine: well people wanted to discuss how we should do things. then it's only fair that people then are allowed to share their ideas.... perhaps some people are just fine with the status quo, they can mention that. Others might want to start their own foundation, and that is fine too. —TheDJ (Not WMF) (talkcontribs) 17:14, 22 August 2014 (UTC)Reply

For whomever posted the Technology Committee idea:[edit]

Hi there, I'm not finding the diff for whomever wrote the Technology Committee section so I'm not completely certain who to address. Can someone please fill it out in the same format as the rest of the ideas on this page? As is it's out of scope. Thank you! -Rdicerb (WMF) (talk) 16:14, 22 August 2014 (UTC)Reply

That would be User:Pine. [3] (Well originally Erik at [4], but the tech committee thing is generally Pine's idea) Bawolff (talk) 16:29, 22 August 2014 (UTC)Reply
I have expanded the section. Note that an idea that is formatted differently or stated more briefly doesn't mean that it is out of scope. --Pine 19:58, 22 August 2014 (UTC)Reply

Hello, a joint committee might be a good solution. I would not be happy with a permanent solution, in which only the WMF decides about software, or only a community/the communities takes that decision. Imagine a committee similar to the FDC, with one part nominated by the WMF, one part elected by the communities, and both parts electing a third part to become complete.

Such a committee, in my humble opinion, would only make sence if the commitee can take real decisions about whether to implement a piece of software. Those decisions must be followed by the WMF, and accepted by the communities. So if the committee decides that the current Visual Editor is not ready to be implemented for all/logged-in/logged-out users, then the WMF must not implement it. If the committee decides that the current Visual Editor is ready, then a community should not come up with another voting that it still rejects the feature. (It might be difficult to guarantee the latter.)

I would be careful to give such a committee too many tasks, such as setting up the whole technology strategy of the WMF or the movement. Reasonably you might have every one or two months a committee resolution about all WMF requests made in a specific period. Ziko (talk) 17:10, 22 August 2014 (UTC)Reply

Prioritization process section[edit]

The Prioritization process was moved for being out of scope. I've clarified it and would like to move it back, as I think there is potential for collaboration and communication with communities about *when* software is developed, and whether we work on Product A or B first (or at all). I think it's a good way of collaborating and working with different communities so everyone understands the reasoning for timing of products. Will move this back into the main list in a little while unless there's reasonably more clarification needed. -Rdicerb (WMF) (talk) 17:29, 22 August 2014 (UTC)Reply

You need a public product backlog. — Jeblad 18:37, 22 August 2014 (UTC)Reply
Jeblad - Thanks for the recommendation. I have seen enough lists on MediaWiki that I thought there was already on in existence. I will look around and link into this section if I find something. -Rdicerb (WMF) (talk) 22:10, 24 August 2014 (UTC)Reply
This got a little out of context, perhaps parts of it should be put in separate sections.
Perhaps more important than when something is developed is why it is developed. You first decide if you really want something, then you decide when you can get it.
Often the SCRUM backlog exist on projects such as this, but the product backlog does not. There are also quite often bug reports for the products, but they often focus on bugs and not feature requests and if they do they are usually to specific to be really useful for planning. There should be some high level docs that says why we want a specific product or functionality, and how it should be built or aquired. Such docs should be written in a non-tecnical language so ordinary users can read them, but they can refer to other technical docs. Do not keep the product backlog as a private document, it must be available for everyone.
Each of the products should have its own "home" somewhere, I think a page on meta is the correct place, and from that page it should be possible to at least navigate to the product backlog, the SCRUM backlog, a rollout plan and a test plan. Basically all the docs that says something about why we want something and 'when it should be delivered.
It should also link to a page identifying the stakeholders and links to the persons in the stakeholder group, and also other people involved in the project. That is, the page should identify everyone that can have an impact on the why and when. — Jeblad 01:20, 25 August 2014 (UTC)Reply
  • This proposal reads as WMF deciding the pririties and then communicating them to the users. Is that deliberate, or did you mean to include a step where the users are consulted or engaged in the prioritisation process? Deltahedron (talk) 06:45, 25 August 2014 (UTC)Reply

Stop hiring people from outside[edit]

Does anyone have any numbers on how many of our tech and design related staff is/used to be a wikipedian/wikimedian before joining the foundation and what positions they occupy ? I'd be interested in those numbers ? —TheDJ (Not WMF) (talkcontribs) 17:49, 22 August 2014 (UTC)Reply

  • I don't think we should presume that the "insiders" would need to already be functioning in a technical capacity on WMF projects (which seems to be the implication of the current wording). For example, I'm in here as a non-technical contributor (mostly to Commons these days, though also moderately active on en-wiki) but I have a Masters' in Computer Science and 30+ years experience in software development. I would venture to say I'm (among other things) a top 2% JavaScript dev and a very experienced TPM. I imagine there are dozens if not hundreds of other editors with similar backgrounds. - Jmabel (talk) 18:57, 22 August 2014 (UTC)Reply
I believe that it's about 50/50 TheDJ, but that number has only been told to me; I've asked HR if any numbers or slides are available. -Rdicerb (WMF) (talk) 00:53, 23 August 2014 (UTC)Reply
FWIW, I was not a Wikimedian before I was hired, except for some baby giraffe edits to law articles and some very early-history edits when I was a stupid kid. Many of our most brilliant employees have similar or even less involved histories. As much as we're happy to hire qualified community members (at least in my experience), I think it would be a huge mistake to stop hiring other people, because we get a lot of quality employees from the greater tech scene. --MarkTraceur (WMF) (talk) 16:02, 25 August 2014 (UTC)Reply
Hmm, their used to be such stats in the monthly wmf reports (e.g. Wikimedia_Foundation_Report,_March_2011#Statistics), but I don't see them in newer reports. FWIW, I don't think hiring exclusively in the community is a good idea. If someone has the skills and is the best person to do whatever the job is, then they should be the one who is hired. However I would hope that existing experience with the community would be considered a major asset when evaluating somebody's resume. I would also hope that new hires without wiki experience receive training on how to use a wiki (or are told to figure out how to use a wiki), along with a basic introduction to wiki-culture. Popular rumor is that the majority (but probably not all) of volunteer contributors to mediawiki who have been contributing over the long term and are not currently in school have received job offers at one point or another (It would be interesting to see some stats on that, but I imagine that's not something that would be releasable). As Jmabel says there are probably lots of potential hires who aren't active contribs to, but I'm not sure how WMF's HR department would figure out they exist unless they apply for a job. Bawolff (talk) 04:23, 26 August 2014 (UTC)Reply
Yes, I wasn't assuming that you would magically identify us, just that active participation in the community, including in non-technical capacities, should be a positive (or possibly even a requirement) for hiring. I gather we have been asked not to edit the content page directly. What I want is an edit to the section "Stop hiring people from outside" to indicate that prior participation in the Wikimedia community need not have been in a technical capacity. - Jmabel (talk) 15:29, 26 August 2014 (UTC)Reply
Also, there could be a broad announcement (rather than targeted to individuals) that you are looking for people who are already members of the community and who have the relevant technical background and would be interested in being employed in a technical capacity.
My sense of the primary reason the community wants the Foundation to stop hiring from the outside is that outsiders don't have a sense for what the experience as a contributor is. For example, my criticism of Visual Editor was that it tried to help users do things they could already do with minimal training -- links, bolding, italics, etc. -- while failing to help with the areas that are difficult -- such as handling tables & images. (If VE could make creating & editing tables much easier, I would have been a very happy adopter; trying to revise a table at w:Eponymous archon while preserving the alternating rows of white & grey boxes was a difficult task that would discourage any new editor.) The folks doing the work of contributing & managing content (as well as proving support to the community by attending to administrative tasks) understand our exact needs from long experience; writing articles is not all about cut-n-pasting text from other Internet sites. People from outside (& those who are away from it for many months or years) often have an incorrect impression of how things get done, & end up giving us tools we either don't need or don't work in the ways they need to. -- Llywrch (talk) 19:22, 28 August 2014 (UTC)Reply
My view is that it's easier to take a great developer and give them a feel for how the projects work than the other way round. One way of achieving that might be for the developers to talk and listen to the community early rather than late. Deltahedron (talk) 19:32, 28 August 2014 (UTC)Reply
I was simply describing what I perceive is the problem. It would be nice if the developers actually talked to the volunteers in the projects & understood their needs, rather than simply make guesses based on what people at the Foundation assume the volunteers need. (But TPTB at the Foundation have assured us that the lurkers support them in email.) -- Llywrch (talk) 16:51, 29 August 2014 (UTC)Reply
It doesn't make sense to stop hiring "outsiders" because realistically the WMF needs the best-qualified person for each job whether or not they have ever written an article or uploaded a photo. However, I think all WMF staff should be encouraged (but not forced) to participate in one or more content wikis such as Wikipedia so they can become accustomed to some of the nuances of Wikimedians. I know that many of them do participate in some way but it doesn't hurt to break down the divide between staff and Wiki-geeks. I appreciate that some staff want to be anonymous and work in the background somewhere but looking at the foundation wiki list of staff I can't help noticing that more than 30 of them are wearing "cloaks of invisibility", one person thinks they are a tiger, and one appears to have a ghostly companion. Seriously you guys and gals need to get out and see daylight more often. :P Green Giant (talk) 07:29, 31 August 2014 (UTC)Reply
The "cloak of invisibility" is the joke name for the default photo used by the user page template, if no image is added. For the most part, people who are "invisible" don't work in the main office and therefore aren't able to get official staff photos made. Whatamidoing (WMF) (talk) 23:16, 31 August 2014 (UTC)Reply

Since we're on the subject, here is the current list of job openings. Jmabel, I think a "top 2% JavaScript dev" might be exactly what they're looking for in this one, so if you're in the market, please apply. Also, if you know anyone who has experience with ContentEditable and is looking for a job, then please ask them to apply as well. Whatamidoing (WMF) (talk) 23:16, 31 August 2014 (UTC)Reply


The normal way to do this kind of development is by creating a stakeholder group where at least one member is representing the user community, and the stakeholder group is free to use whatever means they find proper to acquire knowledge about the problem at hand. The stakeholder group will, if they find it necessary, run focus groups and especially they will have one or more user groups. It is also common to involve a group of users in final acceptance tests, but final in this case can be partial deliveries like some subset of functionality. One from the stakeholder group will act as product owner against the scrum teams, note that this is a product owner not a project owner. At the contractor or subcontractor level there might be project owners, but those are internal for each firm.

I would say use a well-defined and known methodology, one that has a firm idea about the existence of a user community. Don't try to make some ad hoc wiki-we-know-best project methodology, use something that is proven and works together with agile development. Take a look at PRINCE2 for example. — Jeblad 18:25, 22 August 2014 (UTC)Reply

Jeblad - in this case community stakeholders within the product teams are the Community Liaisons. There were user roundtables and research was done for Media Viewer's development process and we're looking at ways of doing all of this more effectively. Thank you for reiterating that this is, however, a key form of engagement. -Rdicerb (WMF) (talk) 22:20, 24 August 2014 (UTC)Reply

The community liaisons would definitely fit in a role as a stakeholder, but then as employees by WMF that has as their duties to communicate the changes. They are unfortunately not representatives for the users as those would be some dedicated people selected from the communities. If the communities wish to be represented by the WMFs community liaisons they should be free to do that, but WMF should not decide singlehanded on this. Actually I don't think it would be wise to do that, but hey, the communities do a lot of strange things.
Usually a stakeholder group will be 5-7 people from various parts of the organization and from users of the product. Use one stakeholder group for each product or product family, but try to limit the scope so they can deliver tangible results. — Jeblad 01:34, 25 August 2014 (UTC)Reply
This is captured as an idea in Technology Committees, Product Working Group, and Per Project Workgroups - I think the second and third could be merged, the first is a formal proposal. Do any of these match what you're thinking, or could you perhaps add to any of these? (I will probably merge the second and third ones, actually, as they are quite similar and both internal WMF-staffer written). --Rdicerb (WMF) (talk) 23:23, 2 September 2014 (UTC)Reply

local consensus[edit]

The current idea of local consensus reads: Get local consensus for each config change or extension installation. (emphasis by me)
I feel, consensus is a conditio sine qua non for the long term success of wikipedia. But it is ridiculous to look for consensus for each and every change. Instead, consensus should be affirmed for major changes.
So I'd swap "each" for "major". Should I introduce a new paragraph? Or should I modify the existing idea?---<(kmk)>- (talk) 22:12, 22 August 2014 (UTC)littleReply

This is about config change, not ordinary page edits. I don't think config changes should ever be done without following some process. There should also be some kind of CMDB, but if all changes refers to tickets in bugzilla that should not be to much of a problem. In addition there should be some kind of incident tracking tool, but that to could be handled in bugzilla. Then there is the discussion about replacing bugzilla. — Jeblad 01:38, 25 August 2014 (UTC)Reply

Is complexity killing us?[edit]


Sorry for my partly broken English, writing it as it comes...

I'm a 9-year contributor (first in 2005 in oc.wikipedia where I'm a sysop now) and my top concern is the technical complexity of the Wikimedia projects. This complexity is an issue as it acts on the contrary of the expected goal.

Please correct me if I'm wrong, but I would like to share these 2 statements:

  • The Wikimedia projects goal is the sharing of knowledge, in a volunteer and community-controlled way - the more knowledge you share, the easier you can create/edit articles, the best it is.
  • The ultimate goal of software is to save all of us time, as we are volunteers, we have a job, a family, some other interests, too, the need for walking and breathing with no screen (even this annoying little pad) to glance at.

Back to software, this means the tools are easy to use, are fit for current disabilities, too, and are implemented / rolled-out in a savvy way.

My experience with software development is now quite sort of a cold case, but I'm afraid self-proclaimed "specialists" have taken over large bits of the wm projects. Even in little projects with 10 really active users like the oc.wp, we've got a user that developed a bot, did not ever bother to run tests nor to ask the community, and massively created articles with wrong links, extensive use of huge templates, poor graphics, improbable colours. This prompted an intervention from outside the community, and the external sysop that started without notice the mass deletion of articles was sort of the same species: "I'm sure I'm a competent programmer, I'm invested by the deity- community, I'll nuke you" (or equivalent diplomatic terms).

I'm aware of communication challenges in a multilingual community where some wikis have 1 sysop and others over 1,000. My onbly recommendation would be to just try to make the life of mere contributors easier. This can be done on two sides:

  • providing a simpler interface for modification and creation (the VisualEditor is a very good start);
  • providing an easier path to understanding the processes in the WM projects, and preventing abuses and misuses.

Best regards,

--Jfblanc (talk) 08:14, 23 August 2014 (UTC)Reply

Adieu Joan Francés ! Vaicí mon opinion personala: Much has been said about Wikipedia's dearth of new contributors, and various technological developments have been justified on that basis, such as the VisualEditor that you (shouldn't!) mention. However, if one reads comments elsewhere on the internet--forums, twitter, etc., the most common if not only complaint I have come across is not about technological barriers (wiki syntax, etc.), but about the attitude of a number of editors and admins, which is often perceived as unwelcoming if not downright hostile to newcomers and less seasoned contributors. It is a bit the elephant in the room that nobody on Wikipedia itself (much less the WMF) seem to be discussing. A Google search for something along the lines of "why I don't edit Wikipedia" will yield some interesting results (a couple of examples: [5], [6], or of course, a whole site devoted to the issue). The logical conclusion would appear to be that less energy should be devoted to attempting a technological fix, and more to changing Wikipedia's organisational culture and then going on a hearts and minds campaign to attract new blood.
The above commentary in no way invalidates your very pertinent points about the technological aspects, but is merely intended to point out that the main killer might be elsewhere.
The preceding unsigned comment was added by (talk • contribs) 20:49, 23 August 2014 (UTC)
Wikipedia has more social problems than most of the other wikimedia sister projects. There are several reasons for that, but one of them is simply that Wikipedia has a much broader mandate that most of its sisters, so the folks at Wikipedia aren't all trying to do the same thing — they're often trying to do different hings that conflict with each other, creating disputes. The other sisters have less of that problem because they are doing more specialized tasks — for example, writing books (Wikibooks), writing a dictionary (Witionary), writing news articles (Wikinews), assembling collections of quotations by authors or on particular topics (Wikiquote). Unfortunately, because those tasks are more specialized, they are also more technically challenging. That problem could be overcome but needs software (I'm working on some, myself) — unlike the problems of Wikipedia, which won't respond to throwing software tools at them. And of course, WMF puts its software development effort into Wikipedia, where software won't help. --Pi zero (talk) 00:08, 24 August 2014 (UTC)Reply

Questions and comments on things that could be taken in considering in the design of this process[edit]

Hi! I didn't find a section where the issue on deployments for different types of users has been discussed, so I'm starting a new topic here. I'm not a tech person, nor am I a super active editor/admin or anything, but in the past two years I've very closely observed some deployments as an editor and as a former WMF contractor and I feel I get a bit of both sides concerns/perspectives in this promising discussion.

  • I understand many current software developments have as ultimate goals to encourage new editors to join Wikimedia projects, by offering a more "up to date"/"easy to use" (in the absence of better expressions coming to mind now) technology;
  • I understand experienced users often see NO advantage at all in new products because they've been using current ones for years and and that never stopped them from editing. Then why should they stop someone else?
  • I understand from the letter addressed to WMF regarding mediaviewer deployment and protection rights that subscribers are requesting WMF to step back on the roll out for all users (no matter whether experienced wikimedians, new wikimedians or non logged-in wikimedians); though I also understand that one of the reasons behind that is because it would impact not only individual experience on visualization, but also copyright information display;
  • I'm not seeing so far discrimination between products that may have high impact on current existing community routines x products with low impact on current existing community routines (considering all will certainly have some level of impact):
  • Even in major technology companies I can not remember of a new product deployment that did not have any bug;

Taking that into consideration, I would like to ask some questions and make some comments/suggestions:

  1. Do you believe that, for designing this new process, it should be taken into consideration whether the new product is deployed for non logged-in users and/or can be easily removed by opt-out and those products that are rolled out to anyone with no opt-out option?
  2. Do you believe that, for designing this new process, there should be any discrimination in the decision-making process for new technology deployments between highly impacting products on current existing community routines x products with low impact on current existing community routines?
  3. Considering that there will always be some level of impact in the dynamics (let's say "increase of participation of new editors still in the intense learning curve period), do you believe it's fair to agree that when the product is really targeting new users, that would be a good reason to live with some temporary issues during deployment until new technology is mostly settled?;
  4. Instead of or additionally to a technology committee, is there already or has it ever been considered having a status for editors to volunteer to test products for a while (I know we can do that by changing beta preferences, but what I'm talking about is slightly different as it wouldn't require us to check new boxes for each deployment)? The idea would have to have a large group of editors cross wikis volunteering to have any new products deployed to them in the first place, allowing a large sample of experienced or new users to report bugs. Like admins, those editors would be key during deployment phases; it could also be required that at least a certain percentage of admins joined this group (as once deployed they might the most qualified to complain and point out problems raised, so at least a few would be required during the test phases;
  5. I would encourage everyone to try to look through non editors lenses and think about why so many new accounts never edited just a bit, or try to better understand why some people start editing and leave, or even why some don't even try editing. Some thorough research/survey on that would be helpful too;
  6. Considering that even in major technology companies new deployments still carry bugs which are only sorted out after deployment, as it's hard to have any perfect product before a long time running it, is it reasonable to agree in this proposed process a certain limited time to have products highly improved after their deployments? Let's say 15 days or something like that (I wouldn't know how much time would be required, but something that can be taken as acceptable)? This would mean we would have an agreement on some level of patience from our editors side, but also WMF tech team commitment to address as quickly as possible the main bugs. If by the time "truce period" is completed, highly priority bugs are still unsolved, then it might mean the best thing to do would be undo the roll-out, work on solving the bugs again in a development environment, and then re-deploy when bugs are solved.

--Oona (talk) 16:57, 23 August 2014 (UTC)Reply

No one asks for perfection before deployment, but I think being essentially bug-free before being made the default experience isn't too much to ask.Kww (talk) 03:28, 25 August 2014 (UTC)Reply
Oona, at the top of Beta Features, there is a box labeled "Automatically enable all new beta features". You can choose this to be opted in to new Beta Features. Whatamidoing (WMF) (talk) 23:31, 31 August 2014 (UTC)Reply
Whatamidoing (WMF), I know. I just think this doesn't make us part of a group committed to testing and giving feedback. Can you tell who opted-in all beta futures to reach out to those people and ask for their explicit advice/feedback on the roll out process like you can reach out to admins when you need to?--Oona (talk) 02:53, 1 September 2014 (UTC)Reply
I believe it would be possible. However, the information is not public, so people might feel like it invaded their privacy to get a talk page message that says "I see you opted in to this." Whatamidoing (WMF) (talk) 23:54, 1 September 2014 (UTC)Reply

"Out of scope" sections in need of review/explanation[edit]

This would be a good place to discuss sections that are out-of-scope for moving or clarification. Perhaps discuss here for explanations of moving or requests for clarification: -Rdicerb (WMF) (talk) 21:37, 24 August 2014 (UTC)Reply

  • That's an interesting way to read it. How does "only hire people from the communities" not describe a process that would lead to greater inclusion of the communities? Kww (talk) 03:26, 25 August 2014 (UTC)Reply
It's more of an HR mandate, rather than a community engagement process that my team would conceivably be able to work with communities on. That being said, it will be parked in the "Out of Scope" on the page. --Rdicerb (WMF) (talk) 19:09, 2 September 2014 (UTC)Reply
As a side note - I also think it depends on the department. For example, the Community Liaisons within Community Engagement will probably always be more effective if they are hired from within the communities!--Rdicerb (WMF) (talk) 19:12, 2 September 2014 (UTC)Reply

Structuring an rfc[edit]

Searchability & Visibility of topic tags in Flow is a related discussion. --Gryllida 05:45, 25 August 2014 (UTC)Reply

"Rebuild WMF as part of the free/open-source movement"[edit]

I read this and it seems to be assuming that the WMF isn't already part of the free/open-source movement. But it doesn't in any way explain how or why the author is assuming it isn't. Anomie (talk) 14:03, 25 August 2014 (UTC)Reply

Is there some useful sense in which WMF is part of the movement? It doesn't look like it atm. In theory it's got some intersection with the movement, but in practice the non-intersecting part is evidently ascendant, resulting in the current crisis (which the WMF shows no signs of caring about, a compelling sign they're not effectively part of the movement). --Pi zero (talk) 14:30, 25 August 2014 (UTC)Reply
I mean at the working level. Put every project into something like sourceforge, in other words, an existing FLOSS framework which already exists and removes the need for WMF to duplicate a layer of management, and allows everyone else in the world to see what's happening, help, advise and contribute. There are a number of existing FLOSS structures and communities, why not put the development work into one or more of them? Forgot to sign Deltahedron (talk) 16:23, 25 August 2014 (UTC)Reply
Note, we used to be an sf project once upon a time. I'm honestly quite doubtful how doing something like that would actually help things (What features in particular do you think, etc would provide that would be helpful compared to the existing framework). There are many large open source projects that are not part of github,, etc. Its not like we're unique in that regard. Bawolff (talk) 18:21, 25 August 2014 (UTC)Reply
I agree. Why do you consider WMF to not be part of the "free/open source movement" just because they host their own repository? In what way does the Gerrit/Bugzilla stack (or the upcoming Phabricator stack) not allow everyone else in the world to see what's happening, help, advise, and contribute? Anomie (talk) 13:31, 26 August 2014 (UTC)Reply
I mean a root-and-branch transfer: doing all the work through existing FLOSS structures. Pu every single development project into open source. Dismantle as much of the internal management mechanisms as you can. Decentralise the staff as much as you can. Open up as much as you can. The advantages would be: cost saving by reducing overheads; increased transparency by getting closer to the volunteer movement; increased use of inrerested volunteer effort; increased credibility and attractiveness to the FLOSS community. Yes, it's radical. Deltahedron (talk) 16:52, 26 August 2014 (UTC)Reply
So most of it is disbanding the San Francisco office? Perhaps you mean something like Your team should work like an open source project or the other link by Erik in [7]? --Nemo 20:45, 26 August 2014 (UTC)Reply
No, I mean it should be an open-source project (or several). There would be a small central team left. Remember, I'm not advocating this as the one and only solution to all our ills. This is an idea for brain-storming. Deltahedron (talk) 20:57, 26 August 2014 (UTC)Reply
You still haven't specified any way in which WMF isn't developing things as open source, isn't using any interested volunteers, and isn't credible or attractive to "the FLOSS community". You've just continued to assert that things should somehow "change" to be "more open source". And you seem to have this idea that a project can't "really" be open source unless no one is paid to work on it; while "no paid employees" might be a valid topic for discussion, it should be discussed openly rather than hiding it behind a wrong-headed definition of "open source". Anomie (talk) 13:37, 27 August 2014 (UTC)Reply
"Open source" is not the same as "open source community" or "movement": one is a technical term about licensing, the other is a group of people. The WMF is not an open source community in the sense I mean it because it is a corporation within its own disjoint management structure which allocates its own internal work projects and priorities rather than having them driven by the wider community. It does not use interested volunteers effectively across all projects, because it does not even know who those volunteers are, let alone engage with them. The contributors to de.wp and en.wp are part of the FLOSS community and WMF is rapdidly losing credibility with them. Further, I have not said, explicitly or implicitly, that open source implies no paid employees, and of course don't believe anyonething of the kind, so that's not relevant.
This is a brain storm -- instead of picking holes in other people ideas, think of some better ones! Deltahedron (talk) 16:27, 27 August 2014 (UTC)Reply
I'm not seeing your point, Pi zero. What are these "non-intersection parts" you're referring to? Or are you getting into social issues of the running of the WMF wikis rather than the development of software? Anomie (talk) 13:31, 26 August 2014 (UTC)Reply
I have to agree with Pi zero here, we have a lot of advocates and free software developers who happen to be working here, and we also happen to be creating free software, but for the most part the Foundation feels like it's not very much integrated with the grander system. The platform department is working really hard to fix that perception, by working with Facebook on HHVM and Phabricator; by working with OpenStack on Gerrit, Jenkins, and Zuul; by working with jQuery's community to some extent on our use of their products, and several other smaller projects besides. But our developers (maybe in part due to some of the observations from this section above) don't always seem to have the best interests of the movement in mind. --MarkTraceur (WMF) (talk) 16:09, 25 August 2014 (UTC)Reply
Operations and Language Engineering are also active in that regard (as well as volunteers), according to mw:Upstream projects#Components (which is missing some of the things you mentioned). --Nemo 16:48, 25 August 2014 (UTC)Reply
There are lots of types of free software/open source projects. MediaWiki is certainly one type. However its mostly controlled by a single semi-hierarchical entity that's concentrating on a single project (And a couple people interested in their own thing, but they are significantly outnumbered). Some people I think associate a "true" open source project as one that is driven by several diverse interest groups or the "community" (However one defines that), with no single interest group in control like how the WMF is. For example, prioritization is often done in semi-private and there is an utterly astounding number of private mailing lists for development teams (And even then, half the time they aren't used, but instead people decide to cc 200 odd people onto an email). There is certainly some benefit to doing things this way, there's also some draw backs, one of them is that outside interests find things much more opaque. Of course, this is not a new discussion, you can find arguments about that going at least back to 2010. Bawolff (talk) 18:21, 25 August 2014 (UTC)Reply
I disagree with your wording, MarkTraceur. As I see it (from the inside), Platform isn't "working really hard to fix that perception". We're not trying to fix some perception at all. We work with HHVM and Phabricator and Elasticsearch and so on because we will be using those products, chosen in part (and no small part) because they're open source so we can improve them where improvement is needed for our use, and we contribute the changes back again because that's what we do as free software developers. BJorsch (WMF) (talk) 13:42, 26 August 2014 (UTC)Reply
  • Someone just wrote on wikitech-l that «As mentioned, MediaWiki is not an open source community at the moment. It is an engineering organization that just happens to have open source contributors». I'm not sure what this statement actually means, but it seems to resonate with some of the feelings others have above. --Nemo 16:36, 16 January 2015 (UTC)Reply
  • Hm. «user/developer participation, service quality, internal and external network size, and quality of external network have been found to reliably predict survival at the initial stage, a weak showing in these measures alerts early signs of trouble». [8] --Nemo 17:36, 17 January 2015 (UTC)Reply

Write (some of) software in wiki markup[edit]

Do you have concrete example of what this paragraph could refer to ? Some example of software use which eventually could be enabled. Because it is a little bit vague as stated. --Xavier Combelle (talk) 20:11, 25 August 2014 (UTC)Reply

What is the next step?[edit]

Sorry if I'm missing something but so far we have 28 process ideas and I have no idea what is the next step. I believe some process are very good ideas and maybe with some clarifications (e.g. "Abandon internal communication" should be specified with examples?) could be taken to a next level. Regards, OTAVIO1981 (talk) 20:33, 26 August 2014 (UTC)Reply

Thank you for this, OTAVIO1981 - this is intended to be an open-ended brainstorm for communities to tell us how we (by "we" I mean everyone involved in Wikimedia projects, whether contributor, volunteer, staff, and even reader) can collaborate more effectively where software development is concerned. Right now the Community Engagement team is currently working with users on the Media Viewer consultation, so this page is a little slower at the moment. It's our intention to have a more structured discussion around all aspect of collaboration throughout the product process, from inception to deployment, in coming months. --Rdicerb (WMF) (talk) 03:14, 31 August 2014 (UTC)Reply
I'm looking forward to help with some ideas that's worth to improve. Thanks!OTAVIO1981 (talk) 19:30, 31 August 2014 (UTC)Reply

Stick to the protocol?[edit]

I'm trying to follow what is going on and I'm guessing part of the problem is that WMF did something unexpected. Superprotect at de.commons.js could be the right thing to do, but it was done fastily by staff without any serious justification (for example, security reasons). I believe there's two problems here: one is concerned to all community and is related to engaging people with different timing of volunteering. WMF's staff virtually have 100% of time dedicated to developing software so if there's little space to engage editors with 1% of time. In general, other process of wikimedia movement suffer with the same problem. The other problem I'm noticing here is about the "unexpected" part mentioned before. am I right? It was really unexpected by tech users and it was implemented quickly? If so, a simple solution could be create a protocol detailing each step of WMF procedure deploys with how many time they could wait for community engaging. Personally, I don't like magical numbers (wait for X day before deploy an important patch) but I have no ideas how to avoid this in order to engage more people in decision-making process. Regards, OTAVIO1981 (talk) 21:01, 26 August 2014 (UTC)Reply

maybe because it was unexpected. But furthermore it was not necassary; thats not the frist wheelwar ever ocured; why so hasty?
and if there is need for superman-rights at all; why WMF-staff? Because they know what's right as they are superior brilliant? - Give it to people elected by the community and who are part of it. E.g. stewards. Its not allowed for sysops/stewards aso. to use their advanced rights to enforce personal opinion. So WMF should stick to that rule as well. ...Sicherlich Post 09:19, 27 August 2014 (UTC)Reply

MediaWiki is free software; open-source misses the point[edit]

Quoting Open-source misses the point by Richard Stallman,

When we call software “free,” we mean that it respects the users' essential freedoms: the freedom to run it, to study and change it, and to redistribute copies with or without changes. This is a matter of freedom, not price, so think of “free speech,” not “free beer.”
However, the obvious meaning for the expression “open source software”—and the one most people seem to think it means—is “You can look at the source code.” That criterion is much weaker than the free software definition, much weaker also than the official definition of open source. It includes many programs that are neither free nor open source.
The New York Times has run an article that stretches the meaning of the term to refer to user beta testing—letting a few users try an early version and give confidential feedback—which proprietary software developers have practiced for decades.

Here, I'd like to see more work toward MediaWiki being free software. Make each user aware of their freedom. Do so by something bigger than "powered by MediaWiki" picture in footer. Free software is a social movement, and it could help with community engagement quite a bit. :-)

--Gryllida 23:53, 28 August 2014 (UTC)Reply

I apologize, my previous message is not to the point. Nothing WMF is doing prevents us from using this software and developing it as we wish. I'm taking that back. --Gryllida 02:48, 1 September 2014 (UTC)Reply

There is always a confusion between MW and WM, Media Wiki (the software) and Wiki Media (the implementation of that software used by the Wikimedia Foundation for Wikipeda and all of the 800 or so projects of Wiki Media, WM). There are potentially millions of projects, no, billions of projects that use the MW software - every website on the planet can include MW for a portion of the site if they choose.
The button at the bottom that says "A WIKIMEDIA Project" has been co-opted by WMF and instead of going to Meta, where it should, goes to the foundation. The foundation needs to learn that they are servants to the community, and not masters of the community. They need to be responsive to the community or go home. I would suggest removing any selection of board of trustee members by the board, and make all of them selected by the community. We also have to thank Jimmy Wales for getting us started, but retire them from having a selection choice of a board member. As all of the Wikimedia projects are "a Wikimedia project", the button needs to go to "Wikimedia" (or Meta), and not to the foundation, as the foundation did not create the projects, the projects created the foundation.
And yes the role of the WMF needs to be more clearly defined. And yes we need to take the "superuser" function away from the foundation, and give it to the Stewards (if it is a Media Wiki function - but if it is a Wikimedia add on, deep six it as a really bad idea). Any "superuser" function, is really the province of Media Wiki, and not Wiki Media. Changes to the software are best proposed and implemented in Media Wiki, so that all of the millions of Wikis can use them, and not have a fork that only our (few hundred) projects has access to use. It is a really bad precedent to have any fork in the software that we developed, and is not available to every other Wiki. Apteva (talk) 15:15, 31 August 2014 (UTC)Reply
I agree that the "A Wikimedia Project" button should link to Meta. --NaBUru38 (talk) 22:28, 1 September 2014 (UTC)Reply
How and where is that done to make that change? Apteva (talk) 14:41, 7 September 2014 (UTC)Reply
I believe that would have to be approved by Legal and Community Advocacy, and be done on the server side. User:Philippe (WMF) would probably know. It is unlikely to be approved as-is, because that's the foundation's trademarked logo, not the community logo. Whatamidoing (WMF) (talk) 05:13, 8 September 2014 (UTC)Reply
You're a bit confused I think. It's clearly acceptable to use the Wikimedia logo to link a domain: there's even a section Trademark policy#policy-linkstous. It's particularly on topic for Meta, of course, because it used to be Meta's logo as well. Technically, it's a trivial change of the $wgCopyrightIcon line of the configuration.[9] --Nemo 07:58, 8 September 2014 (UTC)Reply
A link to Meta should have the Meta logo. It makes no sense to use the Foundation logo (or the Wikipedia logo, or the Wiktionary logo, or any of the other ones) when linking to Meta. This is a widely used and well-understood system for dealing with trademarked logos. Volkswagen owns Automobili Lamborghini, but a link to the Lamborghini website should show the Lamborghini logo, not the Volkswagen one (and vice versa). The logo should reflect the link target, not something sort of related to the link target. Whatamidoing (WMF) (talk) 23:07, 12 September 2014 (UTC)Reply
Sure, this is a legitimate opinion. Nothing more, nothing less. --Nemo 08:15, 14 September 2014 (UTC)Reply

Improve Notification system to upload comments send by mail.[edit]

I'm not sure if this idea is in the scope of this brainstorm process but here what I thought: Yesterday, I got an email that my username was notified here with the comment sent by Rdicerb (WMF) related to a previous topic at this discussion page. What if I answer that e-mail and notification system upload my comment here? I know that facebook system have this system and maybe technology team are already working on this but if not, how good would be have this? At first, more people could be engaged by small comments. Second, I guess it would be easier to follow just what you want. Personally, I don't like to subscribe lists because most of time I'm not interested in a large part of what is posted but sometimes I'm interested to join a particular topic. I think it's also reduces the widespread of information between lists and discussion pages. However, if I'm not watching a particular page I won't receive new topics I might be interested (the same reason I don't subscribe to lists). In other words, 'Beware of the Leopard' could still happens. If someone thinks it's worth to develop such gadget, we could narrow focus to a better shape proposal. Regards, OTAVIO1981 (talk) 13:34, 1 September 2014 (UTC)Reply

User:Quiddity (WMF), is Echo on your list? Whatamidoing (WMF) (talk) 23:56, 1 September 2014 (UTC)Reply
@OTAVIO1981: The reply-by-email idea is not currently being worked on, but has been suggested before. One of the drawbacks, is that the email system only notifies us about specific replies, and hence if someone else in the thread had already answered the question I was about to reply-to-by-email, then I wouldn't know that unless I visited the full thread itself to read anything else that was new.
Re: Second: There is a related idea for mw:Flow (which now enables us to watchlist a single thread/topic), whereby we might eventually have 2 levels of "watchlisting" a discussion page: A) watchlist everything (as usual). B) only tell me about new topics being created (and then I can individually watchlist the topics I'm interested in). -- That should (at least partially) solve the issue of people not wanting to watchlist locations such as VillagePumps, because they're not interested in a majority of topics.
HTH. Quiddity (WMF) (talk) 18:04, 2 September 2014 (UTC)Reply
Thank you for this idea, OTAVIO1981 (as I ping you again :). You mention that you aren't sure if the idea is in scope, and my general thought is that we're asking for process ideas and less feature ideas. That being said, you're highlighting a pain point in that some users, yourself included, do not find list subscriptions useful but still have a need to be informed, and might miss information that's important to you because you don't have your eyes in all places at all times (this is what I understand from your message; do let me know if I've misunderstood you).
On a product level, I mirror my team members in bringing up Echo notifications as a way to think about keeping users more informed, but I think we can expand on that and ask if on a process level, we (and by "we", I mean all communities and the WMF working together) are willing to consider whether more centralized or all-encompassing ways of communicating changes are desired. I happen to see extremely decentralized communications as a problem we can solve in order to keep communities informed - though I think it might be a bit early to try to describe exactly what we need in order to solve that problem. If that is the case, that is definitely withing scope. Let me know if that makes sense (OTAVIO1981, or anyone who happens to read this) - if not, I will try to clarify. --Rdicerb (WMF) (talk) 18:47, 2 September 2014 (UTC)Reply
Quiddity (WMF), I think I understood the drawbacks. However, I was thinking something like that: if notification system send a question to a user by mail, we should only expect an answer to that question. For example: you ping me and ask me:"Does it make sense to you?" and by mail I could answer just "yes, it does!" even if another user like Rdicerb (WMF) have already answered or the topic has advanced in another way. About the problem of watchlist locations like villagepump, at we solved that problem with a gadget that creates a topic as a subpage of the villagepump. By this way I can watch village pump to know just which topic was created and add to my watchlist just topics that I'm interested. Rdicerb, I think about your reply to answer you in the future. But you are right about you understanding and I agree with you in your second paragraph.OTAVIO1981 (talk) 19:41, 3 September 2014 (UTC)Reply
@OTAVIO1981: Hmm, I personally would not want to get any more email (I already get far too much!), but I think I understand how that would be suitable for some people. It's the kind of thing that might be possible with Flow. (More so than with current talkpages, at least.)
Re: centralized vs dispersed communication, and filtering the content of mailing lists or pages: that's another thing that Flow might be able to help with. Some people have suggested using a #Tag system (or possibly the category system), so that we can better Filter topics - both to Filter out topics that we're not interested in, and to Filter in topics that we do want to keep track of. E.g. I might want to get notified of any discussion that has the tags #deletion and #microbiology in it. Maybe even cross-wiki? That's still just at the brainstorming-stage though, and will be for a while.
Re: gadget: Ah! I first noticed that technique at the Italian villagepump, and then learned it is also used at the Danish villagepump. It's helpful to learn that Portuguese uses it. (I'm just linking examples - I do see that it is used on a few more pages :). All 3 seem to use different code-systems though! And of course it doesn't scale very well because of all the setup required, plus it's slightly more confusing for newcomers to learn, hence only the villagepumps are using it (as far as I know?). That's exactly the type of thing that Flow aims to take the good ideas from and further-improve-on, so that all wikis and pages can benefit. (eg. villagepumps and other very-busy talkpages could use the (B) setting, but most quiet talkpages could use the (A) setting.) If you have further ideas about how that might work, now is the perfect time to suggest improvements. (over at mw:Talk:Flow) :)
Hope that helps. Quiddity (WMF) (talk) 02:02, 4 September 2014 (UTC)Reply
The subpage system was made in 2006 inspired by and it was already described in bugzilla:25055. Too bad years of careful requirements collection by the adventurous LQT users have been completely disregarded. --Nemo 04:58, 4 September 2014 (UTC)Reply
Quiddity (WMF), of course some people don't want to receive more e-mail so it means system should be customizable for users. But if users had activated e-mail system to receive updates and pings, it means they don't bother to receive more. Probably, they are not able to access wiki at the moment or it's a wiki (like meta) they don't access too often but they want to be engaged with what is going on. This tag/categorization system looks very interesting. Information and discussion will continue as it is: widespread at N places but it will be possible to know when someone "ping" a topic like "gadget" at any page. Quite often people don't know where to send their ideas or who might be interested to help with but this system will help with that too. I'll try to help with improving of flow system. Regards, OTAVIO1981 (talk) 12:34, 4 September 2014 (UTC)Reply

My opinion[edit]

In a nutshell, I think that the product development process of the Foundation team should be user-oriented. User input should be considered in each step of the development process. The goals of the development process, the development projects and the features of the products should be decided based on what users demand.

I'm speaking about users in general. The Wikimedia community is very diverse, since there are several Wikimedia projects (Wikipedia, Commons, Wikidata), social groups (language, region, income, education, urban-rural, digital literacy) and user groups (reader, writer, admin) involved. So each group should be considered in the development process.

-- NaBUru38 (talk) 22:20, 1 September 2014 (UTC)Reply

Hi NaBUru38, it's definitely a goal to get user input throughout each step of the product process. In each step it might look differently, and we will have to continue to iterate to ensure we're collaborating effectively. --Rdicerb (WMF) (talk) 22:28, 2 September 2014 (UTC)Reply
There are many good ideas within the community. Some help developing these from the WMF would be excellent. It would increase engagement of our most active users. The Foundation should be handing out developer time not just cash. The Foundation is an expert in tech employee hiring. --Doc James (talk · contribs · email) 06:10, 18 September 2014 (UTC)Reply

Make life easier for the editors - generate references in one click[edit]

  • What new tools and processes should we look at ?

Answer: We need a software program and a standard in order to create references in one click.

Creating references is a pain. Creating one well formatted reference can consume between 30 seconds to 1 minute. When you edit Wikipedia a lot and you add a lot of references, you lose hours, days, weeks, months and even years of your life just for formatting references. Sorry to use strong words but this is really insane. Creating references should be just one click away. Editing Wikipedia means doing something useful, and it's a wonderful work. But manually formatting references is not work and it's not exactly editing Wikipedia - it is simply a waste of time. We want a reliable and verifiable Wikipedia, containing a lot of references, so the statements contained in the articles can be verified. I want to add good quality and verifiable information into the article, but many times, wasting my time formatting references is turning me down. And I am an experienced editor. The question always comes into my mind: I am doing voluntary work. I am even sending money to Wikimedia Foundation from time to time. But now I have to waste years of my life for manually doing something that should be done automatically, instead of actually editing Wikipedia in that time or spending that time with my family and friends? I (and other editors like me) have to waste my life simply because some people who can make a decision to fix this easily can't be bothered to think about this? This is really extreme, to use the most beautiful word that I can. And this waste of my (precious - I dare to say) time is turning me down many times.

Now imagine the effect of painstakingly format the references on the newcomers. Most of the time they simply put just the link without any formatting or they even skip adding any reference after the statements they add into the articles. The effects of this situation (from the point of view of the quality or references) is devastating. Check the smaller Wikipedias, most of the references in their articles have junky formatting. Check for example this article, that I edited today: ro:Relațiile dintre Republica Moldova și Uniunea Europeană - only the first three references look decent (and two of them are added by me), the rest of them are poorly formatted. Most of the articles look like that, and even many of the experienced users are adding such poorly formatted references. And I can't blame them: they have families and other things, they can't waste their time with such a nonsense.

If the editors would be able to create references with just one click, then Wikipedia will have much more references, and very well formatted. And this is quite easy to fix.

I made a script that can help editors to create references in one click, but the bad part is that I have to "teach" the script how to handle each newspaper or online publication - because there is no standard for this. And sometimes I have to "teach" it again how to handle a specific site, since they update their software. And it can only handle the websites it "learned", of course. If there will be a (W3C) standard for presenting fields like title, date of publication, name of the author(s) and name of the publication, then the newspapers will use the standard. Then it's easy to create such a script like the one I created, and it will work for any website that implements the standard, without the need to teach the script how to handle every particular website. A newspaper doesn't want to implement the standard? Then boycott it! If the same information is found in USA Today and in Boston Globe, but only USA Today implements the standard, then use USA Today when referencing the information you add into the Wikipedia article. So the newspapers will have to implement the standard if they want to have more visibility through Wikipedia references, which is not exactly a small feature.

How to implement the standard? If the leaders of Wikipedia/Wikimedia will try to reach W3C, I am sure that W3C will listen. The standard can be created in a matter of minutes (something like "<span class="publication date">January 1, 2000</span>"). I am also sure that after a little time, a few newspapers will listen to Wikipedia's and W3C's advice and will implement the standard. And then the script will work with those newspapers. Wikipedia and W3C are not Ark25 - many people listen to them.

You can find my script at en:User:Ark25/RefScript - it's huge and bulky exactly because there is no such standard. But still, it's very useful. The script has very few users and I don't understand why there are so few. Maybe most of the editors can't even imagine how easy their life can be with such a script, or maybe they even abandoned the idea of using references as often as possible.

If you want more Wikipedia editors, then please respect them and don't ask them to waste their life for manually doing things that should be done automatically in one click. I really don't believe that it's so hard for Wikimedia leaders to make W3C listen to them. —  Ark25  (talk) 23:51, 1 September 2014 (UTC)Reply

Will you settle for two clicks? That's what we should get with mw:Citoid. Whatamidoing (WMF) (talk) 23:57, 1 September 2014 (UTC)Reply
Maybe I would settle with two clicks but citoid is not production ready. And it asks the users to install "Zotero's translation server" - really, this is not looking like a user friendly solution at all. The one-click solution to generate references must be a JavaScript program, preferably a bookmarklet (like my script) - no need to install any program or browser extension. —  Ark25  (talk) 03:15, 2 September 2014 (UTC)Reply
In the meantime you could try a couple of gadgets on English Wikipedia such as ProveIt and RefToolbar. Green Giant (talk) 00:31, 2 September 2014 (UTC)Reply
Both ProveIt and RefToolbar ask me to manually fill all the fields: title, author name, publication date etc. RefToolbar even asks me to fill the access date, which is allways today! Thanks for trying to help but this looks extremely far from a one-click solution. My bulky script is already a one click solution and, on Romanian Wikipedia, it "learned" most of the national newspapers. I just click the bookmarklet and BAM! From a link like this I get the formatted reference like <ref name="MyUser_BBC_May_20_2014c">{{cite web |url= |title=Sea otter return boosts ailing seagrass in California |newspaper=BBC |date= August 26, 2013 |author=Suzi Gage |accessdate=May 20, 2014}}</ref>. But the lack of a standard requires me to constantly update the script and to teach it how to handle more websites. And that's consuming time. With a standard the script will be short and no need to update it or to teach it any website. —  Ark25  (talk) 03:15, 2 September 2014 (UTC)Reply
@Ark25: Re: Your bookmarklet: This used to be (many years ago) a popular tool: en:User:Bazzargh/citemark, but it doesn't seem to work anymore. (IIRC, It wasn't great, as it relied on scraping everything potentially relevant and always included a lot of cruft that had to be trimmed out). I mentioned your bookmarklet on the talkpage last year, but possibly we should add a bolder pointer to the top of the project page, pointing to your working version? Or possibly you can fix Bazzargh's bookmarklet as an alternative tool?
In the long-term, Citoid should be exactly what we want (if I understand it correctly). It's currently just at a small-scale testing level, but you can test it out in VE, by using this userscript: en:User:Mvolz/veCiteFromURL. Note that the "Zotero's translation server" is only required for installing the Extension on a server, and not for editors who want to use the tool within Wikipedia. Beyond that, I'm not very familiar with it, but I believe it is doing exactly what you suggest: Encouraging sites to expose their metadata using the existing standards, so that everyone in the world can benefit.
HTH. Quiddity (WMF) (talk) 19:38, 2 September 2014 (UTC)Reply
@Quiddity (WMF): I remember trying to use citemark when your recommended it to me last year. It's a very short script and it was adding bulk of information. If I would improve it, I will have to re-write RefScript. Standards like DOI and programs with Zotero and citoid are designed with libraries' catalogues of books in mind. They are quite obscure and esoteric to the regular editor. We need a very simple W3C standard for the newspapers (not for books or libraries) for publishing just 4 tags: title, date of publication, names of the authors and the name of the publication. A standard for defining tags like "<span class="publication date">January 1, 2000</span>". This is a matter of minutes. We need a tool that creates references based on newspaper articles. I don't know if anyone bothered to count, but I suspect that references from newspapers are in larger number than the references from books. —  Ark25  (talk) 08:31, 3 September 2014 (UTC)Reply
I tried en:User:Mvolz/veCiteFromURL but my VisualEditor hangs now, the blue progress bar in the right corner keeps "progressing" indefinitely. —  Ark25  (talk) 08:35, 3 September 2014 (UTC)Reply
  • It is perhaps relevant to the wider issue here that, as far as I can tell, the old citemark and the new Citoid are both volunteer projects, not WMF staffed or funded. We should perhaps understand how it is such such popular and evidently useful tools are left to the community to develop: what stopped these ideas from being proposed to or taken up by WMF, and what mechanisms would allow such ideas to be captured, assessed and ideally implemented by WMF in the future? Deltahedron (talk) 19:45, 2 September 2014 (UTC)Reply
    @Deltahedron: That's not entirely accurate – Marielle's work on Citoid was done with us under the OPW programme, which is not precisely volunteer-like, and was done at our suggestion and instigation. The broader answer to your question "what stopped these ideas from being proposed to or taken up by WMF" is that we stop ourselves from stepping on the toes of volunteers doing excellent work. It wouldn't be right for us to waltz in and take over unless asked for. Jdforrester (WMF) (talk) 21:12, 2 September 2014 (UTC)Reply
    Thank you for that clarification: as I said, I was unaware of the precise details. Your comment on the broader issue is interesting and I think deserves a detailed examination. How this particular idea was proposed and allocated a place in the OPW programme? Where did the suggestion come from: how did it come about that it was instigated by WMF? Does that mechanism work well, and does it scale? Could it or should it be captured as part of this discussion? As I said, I don't have the answers to those questions. I think this is not the first time the the opinion has been expressed that WMF staff cannot, or do not, or should not, tell volunteers what to do. Well, one possible reaction to that is that in fact they do, and there is, as I'm sure we're all aware, recent rather serious rows about that in the context of Visual Editor and Media Viewer. However, the purpose of this discussion is not to replay those arguments. While it may well be politic for WMF staff not to order volunteers around, there are alternatives to that that are not just masterly inactivity. These proposal need to be coordinated, communicated, combined, and yes, occasionally it may be necessary to tell people not to do something, at least in the sense of telling them that it will never be implemented on WMF kit. The issue is not the simplistic do it or do nothing that is implied here. It is the extent and way in which the two groups of people get together to decide what to do and how to do it. This page is for brainstoriming new ways if doing that. One final point: I am at least cheered by the suggestion that WMF might waltz in and take over when asked. That's a useful concept and worth adding to the list here: WMF as last resort, and the extent to which it needs to act as a long stop standing behind volunteer projects that run into difficulties. Deltahedron (talk) 21:29, 2 September 2014 (UTC)Reply
    [ec] @Deltahedron: Sure – The idea to have "magic" auto-filling citations came up in talks in the Product team between Erik M., Jared Z. and myself about how we could make citations better for all (in VE and more widely), and similar but vaguer wishes (including "please can we make enwiki's awesome RefLinks tool available for all wikis") had been suggested by the community a few times. Once VisualEditor was enabled by default, we suddenly got a lot of feedback (which was great, if a little later than ideal for making VisualEditor something people liked before it was deployed), several of which touched on this area like bug 50110 and bug 50768, and finally bug 60768 (filed by Jared). I prioritised this as the number two priority for VE, following on from 50768 itself (i.e., auto-filled citations should come straight after advanced citation support), and I later suggested it as a good piece of relatively self-contained work for a volunteer to look at if we hadn't got to it yet. Come April, Marielle found our posting, said she was interested, and when we spoke to her we felt it could work well so we approved her to be our OPW student, committing to mentor and support her during the work. This went very well, and we're now in the process of taking Marielle's excellent system and getting it ready for production.
    As to whether this process scales, it's good at a few things (bringing new coder volunteers into our world; getting fresh ideas, technologies and ways of working; spreading the load for important but less urgent that WMF won't get to as quickly), but it's still pretty centralised on WMF staff availability, especially on the mentoring and deployment sides. Jdforrester (WMF) (talk) 18:42, 3 September 2014 (UTC)Reply
    That's interesting, thanks for the details. One thing that strikes me immediately is that a new feature requested by users was requested through Bugzilla. It seems that most users, including myself, would not view a request for a new feature as on the same footing as a bug report; that most users would never have heard of Bugzilla; and that Bugzilla is very badly adapted for use as a forum for discussing, proposing and designing a new feature. That seems a major weakness in the current way of doing things. Is the mw:Mentorship programs/Possible projects available to the user community as a forum for proposing work to volunteers, or is it too far downstream? Deltahedron (talk) 18:49, 3 September 2014 (UTC)Reply
    @Deltahedron: Yeah – Bugzilla, despite the name, is not for bugs, it's for code change suggestions. Everything from "the cluster is down" to "I'd like to beam copies of Wikipedia articles into space". Soon (a few weeks' time) we'll be moving to mw:Phabricator, which is a bit clearer for non-bug management; Bugzilla does indeed have a lot of flaws. Also, yes, everyone is welcome to suggest ideas on mw:Mentorship programs/Possible projects that they'd like to see come to fruitition. However, in practice OPW and GSoC students are required by the sponsoring organisations to have mentors (so, at some point you have to sell it to WMF staff if you want it done through those routes). Quim and the rest of the mw:Engineering Community Team semi-regularly send out reminder e-mails to wikitech-l and wikitech-ambassadors-l, and put entries in the Tech/News newsletter asking for entries from interested parties. Jdforrester (WMF) (talk) 19:03, 3 September 2014 (UTC)Reply
Again, thanks. I was interested in "at some point you have to sell it to WMF staff". Is there a recognised process for doing that? Again, you'll understand we're trying to tease out what is missing in the current way of doing things. Is it possible to get the Tech/News newsletter sent to a WP project talk page? Deltahedron (talk) 19:34, 3 September 2014 (UTC)Reply
Additionally, it seems to be taken for granted that when you say OPW and GSoC students are required by the sponsoring organisations to have mentors, those mentors have to be WMF staff. Is that a formal requirement of the various sponsors, or just an unspoken (until now) assumption. It doesn't seem obvious that the community would be unable to provide a suitable mentor. Deltahedron (talk) 19:40, 3 September 2014 (UTC)Reply
  • @Deltahedron: Some good questions. :-) It isn't always required for the mentor to be WMF staff (and there have been times when that's not been the case), but generally people want to get code into production and that means deployment access, which though not exclusive to WMF staff is very rare outside of that (something we're slowly improving on); additionally, WMF staff can open up conversations, or know to whom to talk, much more readily than almost anyone, so a lot of that side of the burden falls on us. To get Tech/News on your talk page (or on a village pump), you can subscribe at Global_message_delivery/Targets/Tech_ambassadors – enwiki for some reason is subscribed on en:Wikipedia:Tech_news as well as VPT, for example. Jdforrester (WMF) (talk) 22:01, 3 September 2014 (UTC)Reply
Thanks for that. Deltahedron (talk) 06:23, 4 September 2014 (UTC)Reply
@Deltahedron: thanks for your answers, we are on the same page. @Jdforrester (WMF): WMF should create the tools the users need in order to efficiently edit Wikipedia. It is true that the users should have asked for this feature, but probably they can't imagine that this feature can be implemented. Sometimes the users have no idea that certain improvements can be made. Which is where the administration/management should step in: offer them the tools that make their life easier. I don't know how many users asked for the notification system that was implemented relatively recently, but in my case I could not imagine such a feature so I could not ask for it. But WMF implemented it and that's wonderful. The same here. So WMF really has to take over sometimes, and that doesn't imply stepping on editors' toes. The lack of imagination of the editors is not exactly a valid excuse for WMF for not doing such things. —  Ark25  (talk) 07:41, 3 September 2014 (UTC)Reply
Glad to hear it. I think the problem is that writers, editors and readers, even volunteer developers just don't know where to go to ask for something. That's something addressed by some of the proposals here, such as Community engagement for all, which proposes a mandate on every member of WMF staff take ownership of such suggestions, or, at the opposite end of the spectrum, Triangulum which proposes a "one-stop shop", a unique portal for all such requests. Deltahedron (talk) 17:12, 3 September 2014 (UTC)Reply
Ark, you don't need to be able to imagine everything yourself. You only need to be in the same community as a person who can imagine it. So perhaps you couldn't imagine notifications, but someone else did; perhaps that person can't imagine decent support for bibliographic citations, but you did.
Similarly, you don't need to know where to ask for everything yourself. You only need to be in the same community as a person who does. For you and Deltahedron, this means that whenever you have a great idea involving something technical, that you share the idea with the rest of your community at one of the Village Pumps (or some other centralized forum). If the rest of the community is interested, they'll point you to the right place (whether that's a bot request page, Bugzilla, or somewhere else). Whatamidoing (WMF) (talk) 01:32, 4 September 2014 (UTC)Reply
This is frankly just wrong: it doesn't work. My community is that of mathematics writers and editors on Wikipedia, and we share our ideas at en:WT:WPM. I have experience of how know well that worked when we needed to engage with WMF on technical issues. For the benefit of other readers, the executive summary is: it didn't. Certain members of "the rest of the community" were positively and deliberately rude and unhelpful. I find it a quite astonishing assertion from Whatamidoing (WMF), a Community Liaison (Product Development), whose job is to ensure "that our community is represented in the decision-making process and that our planned software adequately reflects user needs". It sounds almost as if your idea of doing that to for us to post to random places and wait for someone to help. That is so far from realistic that I can only assume that I have radically misunderstood your advice. I am going to be so rude as to ask, explicitly, just what it is that a Community Liaison does. Deltahedron (talk) 16:54, 4 September 2014 (UTC)Reply
More relevantly to the topic of this page, perhaps I could ask Whatamidoing (WMF) what suggestions she has for this page. Is it her opinion that the existing method for technical suggestions (1) is simply "share the idea with the rest of your community at one of the Village Pumps (or some other centralized forum)" and (2) works well enough that no changes are necessary or desirable? Or is it possible that we do just a little bit better? Deltahedron (talk) 17:03, 4 September 2014 (UTC)Reply
Actually, the standard, community-based process did work: you found out who and where to ask, and you received an answer. You did not like the answer that you received (namely, that mathematics rendering is, has always been, and for the foreseeable future will continue to be, managed by the volunteer developers who also happen to be members of that same subcommunity of the people who write encyclopedia articles about mathematics and whom you were talking to, instead of by paid staff), but you did receive an accurate and very reliable answer.
I have previously told you exactly what my job entails: supporting exactly one product, and specifically supporting a product that is not mathematics rendering. If it would be helpful, I can find the diff so that you may read it again.
As this is a brainstorming page, and as I will be one of the people reviewing the suggestions in the end, it would be preferable to let others make proposals instead of filling the page with my own ideas. Whatamidoing (WMF) (talk) 18:49, 4 September 2014 (UTC)Reply
How dare you? That is not true, and I think you should withdraw it. I am not complaining here about the answer I got, and it is misleading, unhelpful and thoroughly offensive to suggest that that is what I am doing. I am making a statement about the process. The process did not work, unless you believe that the "standard, community-based process" is or should be, to refer every such request to Jimmy Wales. I did a lot of work with rather little help, and a surprising degree of obstructiveness, from the paid staff whose job I had, perhaps foolishly, thought it was to help, not hinder that. You have told me what your job description is. I am asking you to contribute to the discussion in a productive way by explaining how you do it. Deltahedron (talk) 19:07, 4 September 2014 (UTC)Reply

Sorry for missing two days, I had the feeling that the discussion didn't continue and I was also busy finishing some edits at RO.WP.

@Whatamidoing (WMF): Yes, someone imagined notifications and they asked for it's implementation somewhere, but I could not find the request to implement it. So I could not support the idea. Even at this moment I can't imagine where to start to search from in order to find the initial suggestions to implement the notification system. So I am in the same community as a person who can imagine notifications, but that doesn't help that much. Luckily, others could see the suggestion and they supported it. But, most likely, for every one person who supported the suggestion, there are 20 others who did not but they would have supported it if they would have seen the suggestion. Also, the other editors are in the same community with me but they can't see my suggestion about a faster way to create references - because the current way to support suggestions is quite messy.

Wikipedia has nice means to communicate, like Village Pump and Help Desk and pages like this, but they are far, far from enough. The current ways to make a good suggestion get the visibility and support it deserves is extremely unproductive. I think I have said it before: we need a regular forum (using phpbb or vbulletin or such) dedicated to Wikipedia. Or a forum for all WMF and with a sub-forum for Wikipedia, another for Wikiquote, another for Wikibooks, and so on. A forum will only complement the Village Pump and Help Desk. In case the forum will make people abandon VP and HD, that's for the better. On a forum you can find a suggestion made by someone two months or years ago, and "bump" it (vote for it's implementation). I am quite sure that Village Pump is full of interesting or even brilliant suggestions that are archived and people can't see them and therefore they can't support them. On a forum it's much easier to notice a good suggestion and to support it, because you don't have the time limit of a month to answer in a thread - the conversations are not moved into archives. Moving from bugzilla to Phabricator looks like a big improvement, now it's the time to upgrade conversations too. On a forum, the most important suggestions will be on top and get more and more visibility. Even such brainstorming sessions like this would have much bigger visibility. The best ideas will be on top (in the list of the threads), because the people will tall about them and the forum engine will put them on the top. Please make a forum for Wikipedia!

I know it would ask for a lot of effort, but a forum with a Wiki markup — so you can easily link WMF pages and use all the nice MediaWiki features — would be awesome. Or maybe it's possible to create an extension for PHPBB to do that? By the way: for me making a blog on regular blogging platforms it's a mess. It's consuming a lot of time. Making a blog in a Wiki page makes blogging a pleasure!

@Deltahedron: - I don't know the details of the talks about mathematics you were involved in, but most likely, having those discussion in a forum would have attracted much more feedback, including more positive feedback. Whatamidoing was very nice with me on Wikipedia, she was helping me a lot. What we need is a better way to communicate, and I can't see a better option than a forum. I hope you will support the idea.

I think I'll make a separate section in this page for implementing a forum. We should make the communication much easier and we should give to the good ideas the visibility they deserve, instead of throwing them into the archives bin. Look at the header of this archive: "If you wish to start a new discussion or revive an old one, please do so on the current talk page". Fragmenting suggestions in 10 archives really doesn't help them to become visible. —  Ark25  (talk) 21:30, 6 September 2014 (UTC)Reply

I found the message where I suggested to implement a Wikipedia forum (Village Pump, February 28, 2014). Comically enough, the message was not even archived. So, good luck Wikiepdia editors, trying to find a valuable suggestion on the Wikipedia's centralized forums! How about a Google Search?

Even more comically, I just found that the messages where the archiving problems were reported were also not archived. How can we know how many times such conversations were not archived? —  Ark25  (talk) 22:16, 6 September 2014 (UTC)Reply

Ark, have you tried out mw:Flow yet?
One of the problems with the "one forum to rule them all" approach is volume. You could literally spend 12 hours a day, every day, reading comments, questions, and suggestions that people make. If it's all centralized, from 800+ wikis, the odds are good that most people will give up. With the de-centralized approach, you can usually post once or twice and find someone who knows who to talk to. Whatamidoing (WMF) (talk) 01:06, 7 September 2014 (UTC)Reply
I will try Flow, as soon as I can.
The forum won't get too big. The forum will have sub-forums. A sub-forum for Wikibooks, another one for Wikisource, another one for MediaWiki, another one for EN.WP, another one for DE.WP, and so on. You won't have a single sub-forum for all suggestions. If you want to make a suggestion for improving just WikiSource, you post your message in the WikiSource sub-forum. The most popular suggestions will be on top, that's a huge advantage. —  Ark25  (talk) 16:37, 8 September 2014 (UTC)Reply
@Whatamidoing (WMF): where can I check an example of Flow at work? Or there is no such place yet? —  Ark25  (talk)
Hi User:Ark25,
You can test Flow at mw:Talk:Sandbox. Make sure you're logged in when you get there. I don't know whether it's set up for IP users to test it, and there were a ew reports a while ago about some browsers not getting people logged in correctly when they went to You can test at the sandbox page; feedback should probably be left at mw:Talk:Flow. Whatamidoing (WMF) (talk) 03:39, 13 September 2014 (UTC)Reply
Ok, I got an idea, thanks. The forum is not supposed to replace the talk pages, like Flow does. The forum is supposed to complement the talk pages. And it has a united, intuitive interface, making it easy to find the best place to talk about a specific topic. —  Ark25  (talk) 21:17, 13 September 2014 (UTC)Reply

@Whatamidoing (WMF): Sorry for being a pain, but can you please try to estimate how hard it would be for the WMF board to approach (or call) W3C and try to engage them in a discussion about implementing such a standard for publications? It's just four fields: title, date of publication, authors names and publication name. It should not take long to create such a small standard. Is it hard to approach them? I am asking because I really have no idea how accessible or inaccessible they are. —  Ark25  (talk) 02:41, 16 September 2014 (UTC)Reply

Hi Ark25, i totally agree with you about the need for easy 1-click-references! Actually i think almost everybody does. A few quick comments:

  • I think this topic is not ideally discussed in depth on this page, i'm afraid ideas presented here will be lost in archives ;-). A better place might be mw:VisualEditor/Design/Reference Dialog and its talk page!
  • Re: metadata standard for web publications: Have a look at Zotero, Using an Open Standard for Exposing Metadata.
  • Re: "We need a tool that creates references based on newspaper articles. I don't know if anyone bothered to count, but I suspect that references from newspapers are in larger number than the references from books." I have collect some stats about refs /links in Wikipedia at de:Benutzer:Atlasowa/ref_stats. It's a power law distribution with a very long tail. And indeed, newspaper sites like NewYorkTimes or BBC are at the top, so if we can improve refs for the top50 or top300 newspaper URLs (metadata for title, webpublisher, date,), this has a lot of impact. And for the long tail of refs to less popular websites, we need to scrape available metadata so that users can edit/refine the citation/ref for Wikipedia templates. Have a look at my stats collection, there is even a weblink-stat for romanian WP of 2011 (amongst other. maybe you want to compare to current stats?) and TemplateTiger supports rowiki too.
  • Re: tools for easier ref generation: I collected a list of 20+tools that wikipedians use for easier ref generation: by bookmarklets, firefox addons, toolserver tools, new tools.wmflabs tools, enWP RefToolbar, user scripts, standalone software, ref generation websites etc. See de:Benutzer:Atlasowa/ref_citation_tools, there is your tool too :-) I think there is a lot to be learned from these tools to build a widely used, standard, easy ref generation tool with autofill/-generation by URL for normal Wikipedia users, for everybody. And ISBN, DOI, PMID refs by autofill too, of course :-) --Atlasowa (talk) 08:12, 17 September 2014 (UTC)Reply
Hi Atlasowa, thanks a lot for your answer! The script that generates references works as a bookmarklet, in one click, no matter if you are using Visual Editor or not. In fact, I never tried it with Visual Editor. But if you think it's better to talk about it at mw:Talk:VisualEditor/Design/Reference Dialog, then please move our conversation there, and I will continue this chat in that page.
From what I can understand, Zotero and the similar solutions there are designed for libraries of books, not for newspaper articles, which are much more "frugal" to say so. Therefore, for newspapers we need a extra-lightweight standard, with only 4 or let's say maximum 7 fields like "<span class="publication date">January 1, 2000</span>". Such a standard would be easy to digest and implement by the IT departments of virtually any newspaper. Or it can be a standard with many versions, where the easiest version (let's say it will be called "RefStandard A") will specify only 4 fields (title, authors, publication date and newspaper name). Version B will have 10 fields, and version C will have 50 fields. Really, when I read about COinS and other similar standards, my head hurts - and I have quite a little bit of experience in computing and internet. Those standards are quite complicated, we need something doable in a snap.
I don't really know who should publish such a standard. I was just guessing that W3C should develop it. But also WMF can design such a standard and then talk with a few newspapers, asking them to implement it. I can't believe that all the newspapers would refuse WMF. Wikipedia is a big thing in the internet, I can't believe the newspapers will play deaf when WMF will talk to them. Maybe this can be done as a research program, and some foundation or the governement can donate some funds for implementing this project, although I don't think there would be a need for any significant amount of money.
Nice statistics there at , thanks.
I think I tried all the tools for making references, but they are not quite useful or they are outdated. If I would have found a better tool than the script I had to write, than I would surely not wasted a lot of my time to write and constantly update the script.
Please help me with this endeavor because, really, it looks like almost nobody understands how important is to generate references in one click. Or at least I can't seem to be able to make the script and the issue known to the community. We need that tiny tiny standard and then we need WMF to convince a few newspapers to experiment with implementing the standard. And then the problem is solved, and creating references will not waste huge chunks of editor's time anymore. —  Ark25  (talk) 23:55, 17 September 2014 (UTC)Reply

Current state[edit]

I think these changes [10] altered the meaning. In particular NPOV in all its rigour is probably not an appropriae yardstick to apply here. This is a discussion, not an encyclopaedia article. I agree that " For a few major, high-priority pieces of work selected by Wikimedia Foundation senior management, these are scheduled and resourced with Wikimedia Foundation staff development time. Other changes happen without much change management of what is and is not "allowed" or part of the wider plan, especially for volunteer-led work or changes in poorly-supported areas." probably expressed view of the facts, but that point of view is in istself a reality, and certainly not uncommon: and I agree with it. If it's unpalatable, then so be it. If you want a more diplomatice rendition, you might either preface it by "Some users believe that ..." or reword it to say "Levels of resourcing and change management principles vary widely across work package", but to remove it altogether is to make things appear rosier than they actually are. Deltahedron (talk) 06:15, 4 September 2014 (UTC)Reply

That change was pure POV, sugarcoating the things the WMF has done against the communities with some meaningless blah-blah. Another fine example of the whitewash strategy by the WMF. Why are they acting so aggressive? --♫ Sänger, superputsch must go (talk) 07:52, 4 September 2014 (UTC)Reply

It was a whitewash, and I have restored portions of it. It is a problem that the WMF seems to want to portray this issue as solely a communication problem instead of a behavioural problem. This would seem to imply that they will never correct the root behaviours. There's no doubt that VE was released prematurely: I think even James and Erik finally admitted their mistakes on that one. There's no doubt that when VE was turned to opt-in on English Wikipedia, both of them actively lied about the effects of the patch that disabled it. That kind of behaviour is unacceptable, and no amount of community engagement will correct it.Kww (talk) 14:05, 4 September 2014 (UTC)Reply

I think that Deltahedron is concerned about a different change than you. Whatamidoing (WMF) (talk) 20:30, 4 September 2014 (UTC)Reply

Phillipe, in this statement, James Forrester actively lied, with the apparent approval of the WMF. The patch caused no damage to English Wikipedia. It was not "known-broken": the server load problem had been repaired prior to implementation, because I take pride in my code, test it thoroughly before deployment, and listen to feedback. I specifically invited Catrope to review the repairs. James knew it had been repaired. He's repeated that lie. I can dig around and find places where Erik repeated it. Hell, I wound up listening to people impugn my competence and integrity on Wikipedia Weekly. No one has ever withdrawn the false statements or apologized for making them. As for superprotection, I think the timing and the fact that you left German Wikipedia in a state that ran against their explicit consensus speaks more loudly to the motivation than any words could: superprotect was implemented in order to ensure that MV remained an opt-out feature on German Wikipedia. What kind of citation would you like?

This really does get to the heart of the matter: try for a moment to deal with the concept that the WMF has misbehaved. This isn't a case of it only being a matter of providing better explanations for your behaviour and then everyone will like you again. Part of engagement is a willingness to admit that you have done wrong, and a commitment not do so again in the future. I've stated before and will state again: if the WMF would simply commit that the next time a WMF employee tells bald-faced lies he would be terminated, that would do more to heal the gap between the WMF and the communities than anything. Kww (talk) 00:59, 5 September 2014 (UTC)Reply

What is the limit to customization of mediawiki?[edit]

One concerning of WMF is that more customization of MediaWiki would make really hard to improve the software. If I understood correctly, with more differents configurations, there's a need for more work to a new feature works correctly in all of them. Well, that's make sense but what is the actual state of those different configurations and how they are impacting to editors, readers, newcomers and WMF? It would be great have a report of what is different right now and how each change one impacts the project. Maybe some configurations are outdated and communities could change to default. OTAVIO1981 (talk) 23:30, 4 September 2014 (UTC)Reply

Hello, this is all Meta-Wiki is about. We do that informally about everywhere and durably on Category:Cross-project comparisons (usually for bigger things), where you can add your own for a specific thing of your interest (if it's not Wikimedia-specific, you can also use A complete catalogue is impossible and would be unusable anyway, while specific areas can and are worked on collaboratively. WMF is not equipped to understand, let alone develop, 11 products in ~280 languages (would anyone be?) hence the development and customisation of MediaWiki has always been and will always be, mainly, the community's responsibility. Everyone who knows something peculiar about the usage of MediaWiki close to them should write it down. --Nemo 06:34, 5 September 2014 (UTC)Reply

The sky is orange[edit]

I've been loosely following the various discussions / dramas regarding Visual Editor, Media Viewer and Flow. Much of what needs to be said has been said, so I'll try to avoid repetition.

I think the biggest problem is WMF product teams are so focused on shiny new gadgets they're missing something essential. For example, when Microsoft rolls out "Office 2014," the most important feature will be the ability to read Office 2013 formatted documents. Of course, there's benefit to updating technology to keep with current standards / user expectations and adding functionality, but the first thing that needs to get done is capturing existing functionality; this has implications for both implementation and roll out.

I get and support the concept that WMF can improve the projects by making projects accessible to a greater diversity of editors and readers. But that should mean supplementing the existing editor-user base, not replacing it. The responses by various WMF staffers lead me to conclude one of three things: a. they're malicious folks on a power trip who enjoy torturing project editors, b. they're really not very bright, or c. they simply do not understand existing user culture. Which is understandable, in a sense -- when you're building a development team you want folks who can correctly code new features. But Wikipedias are different than either captive corporate audiences (employees are paid to do a job, no matter how annoying the software is), or competing in a crowded market space (if you get sufficient customers to cover your costs, you're doing well). The goal should be to make Wikipedia software work for all volunteer editors.

After 7 years hanging around en-Wikipedia, I've gotten use to the various changes over the years -- the edit button that used to be "here" is now "there," and the orange talk notification disappeared -- and came back smaller. All annoying, distracting, not terribly important and soon acclimatized to.

But when there are are major changes -- the VE, MV and Flow changes, it's like waking up one day to find the town all proud of their new street signs and I'm like the sky is orange! Street signs are better? Sure, but who cares -- the sky is orange!!!

So, the design implication should be how can you maintain an interface representation that's as close as humanly possible to the existing view for experience users, while providing an optional new interface for new users? For MV, for example, the most important thing that should have gone into the new interface first is a "Switch to classic view" with a "remember this decision" option. While not everyone would have been happy, that would go a long way to satisfying many folks. (To be explicit here, "representation" means both appearance and functionality.)

The rollout implication should be getting all the existing functionality into the new thing before calling it "beta." First impressions are the most important, so imposing barely functional proof of concept toys into the communities is a bad idea. I turned on VE once months ago -- as the adage goes, Fool me once, shame on you, fool me twice, shame on me. I ain't turning it back on as long as I have any choice in the matter. NE Ent (talk) 21:36, 6 September 2014 (UTC)Reply

Well, I wish that you would turn VisualEditor back on, and give it another try, because it has changed a lot. However, you should remember that there's a learning curve to any major software product, and just like you once had to learn that you type double-curly-braces to transclude a template in wikitext, you'll have to learn to click on Insert > Template to do that in VisualEditor.
As for "maintain[ing] an interface representation that's as close as humanly possible to the existing", I'm curious: Are you still using MonoBook for your skin? It's not even close to the original skin, but it's "the original" for when you started editing. Whatamidoing (WMF) (talk) 01:03, 7 September 2014 (UTC)Reply
Say what? I have no interest in Visual Editor no matter what "improvements" are made. If I can not see the mark up that I am editing, I can not edit. Having it WYSIWYG is useless - that is what "show preview" is for. I have no problems with anyone who thinks it is the best thing since sliced bread, but for me it is completely useless, and for people like me the option of not using it always needs to be available. Apteva (talk) 20:06, 7 September 2014 (UTC)Reply
I have to agree with NE Ent and Apteva, I really don't enjoy Visual Editor at all. Wiki markup is easy enough. I also prefer not to use Media Viewer. It might be a good idea to create a site like, in order to address any questions the users have about wiki markup and features. Wikipedia documentation is very good but sometimes it takes too much time to find out how to do a particular thing, when you search it in multi-page documentations. It's a very good thing to have a question-answers feature, makes learning much faster.
Having said that, I think it's good to have VE and MV, they might become powerful tools in the future. —  Ark25  (talk) 11:00, 12 September 2014 (UTC)Reply
Whatamidoing (WMF), I agree that wikipedia changed a lot since I started editing in 2007. I missed a lot the Orange Bar of Doom but I get used to the new one since they do the same thing :). People wouldn't bother so much about MV if they could do the same thing or could switch to the old view for a while. And I think that's why VE didn't create so much problem: we can still edit by the old way. We still can make our contribution without bother to learn new features. It takes time to learn how to do everything we do with wikimarkup using visual editor (e.g. templates, navigation's bar). I feel worried about flow and how WMF is conducting the roll out. It will change dramatically the interface and so far we won't be able to do a lot of things. I'm not saying WMF should limit the software improvents to cosmetic changes at the interface. What I'm saying you should really stretch the process of change from one culture of editing to another. regards, OTAVIO1981 (talk) 12:41, 12 September 2014 (UTC)Reply
People can opt out of Media Viewer. The preference setting has been present from the very beginning. Consequently I think that your statement that "People wouldn't bother so much about MV if they could do the same thing or could switch to the old view" is unfortunately false: people are upset about it even though they can easily opt out whenever they want. Whatamidoing (WMF) (talk) 02:20, 13 September 2014 (UTC)Reply
Sorry, I didn't notice the switch off of MV. But the idea behind "the sky is orange" is still valid. Regards, OTAVIO1981 (talk) 11:56, 15 September 2014 (UTC)Reply
Commons:Category:Orange sunsets? ;) Anomie (talk) 13:28, 8 September 2014 (UTC)Reply
Look at my new pet!
Look at my new pet! Look, cute!
...and if you don't like orange you should get yourself user-preference sunglasses. If you buy the special sunglasses, problem fixed, everything is fine. If nobody buys the special sunglasses [11][12][13], then everybody obviously likes the orange sky. See: there is no problem, either way ^^
Now, look at my new pet! Do you think it needs a cute ribbon? We can discuss color options for the ribbon! --Atlasowa (talk) 09:38, 17 September 2014 (UTC)Reply

Eject the software mission from the WMF[edit]

This can't be fixed and the more people are trying to fix it, the more messy it becomes... I suggest we just close the doors and return the software mission to the community. It is currently doing more harm than good having this under the WMF umbrella. —TheDJ (talkcontribs) 13:09, 7 September 2014 (UTC)Reply

Let me just note that "the software mission" has never been injected in the WMF in the first place. ;-) --Nemo 08:00, 8 September 2014 (UTC)Reply

Community-facing rights and responsibilities of employees document[edit]

I want to suggest that I think it would be supremely useful to have a document that outlines the rights and responsibilities of employees. This would be a community-facing document designed to set some minimum standards of communication for employees, and clarify what the responsibilities of employees are.

Some examples of responsibilities might be:

  • If an employee is monitoring a conversation, use a template to let users know.
  • Respond to communications on monitored pages within 48 hours
  • If unable to answer a query, ping to an employee who can, or refer the user to an appropriate venue.
  • Treat users fairly

Some rights might be:

  • Be treated with respect
  • Refuse to engage in some conversations, including about things said that are off-topic, areas and topics outside the team the manager is working on, and the role-out of features in alpha phase.

Why I think this would be useful is that it's clearly distressing and adversarial for the developers and community to engage in ad-hoc conversations. I think setting some standards for what, when and how developers communicate helps the community, by helping understand what manages do, and helps WMF employees, by offering a document that they can point to when explaining their role and style of communication. This will reduce the current ad-hoc arrangement and, I think, improve and standardise the general level of communication. I was prompted to write this by this thread on [14]. It's not fair for @DannyH (WMF) to have to deal with a large amount of hostility, but then I think the reason for the hostility is because community members don't know who will answer questions, when they will be answered, and whether to expect answers to be relevant or not.

Templates on monitored pages[edit]

It would also be very useful if pages monitored by WMF employees could have a template attached at the top to say that they're being actively monitored by employees, and then the names of the employees responsible for communicating noted. --LT910001 (talk) 22:50, 12 September 2014 (UTC)Reply

Update: I have been 'bold' and created and added the template 'WMF present' to this page and 2 on as a demonstration. --LT910001 (talk) 23:14, 12 September 2014 (UTC)Reply
This is not a bad starting point, but there are some practical issues with the current formulation:
  1. If an employee is monitoring a conversation, use a template to let users know.
    • What constitutes "monitoring"? You know me: I read dozens of pages each day, but I might not get back to the same page for a week, or a month. Is that "monitoring"?
  2. Respond to communications on monitored pages within 48 hours
    • This idea needs some work. Right now, it says, "Employees aren't allowed to go on vacation, get sick, or even have a normal weekend, because if I post a question on Friday evening, I'm entitled to a reply no later than Sunday evening." Also, not all communications benefit from a response from WMF staff, and it conflicts with the proposed right to "Refuse to engage in some conversations".
  3. If unable to answer a query, ping to an employee who can, or refer the user to an appropriate venue.
    • This assumes that the query still needs an answer and that the employee knows how to find the right person or venue. That's probably a safe call with someone like me at en.wp or if the question is about on my current or past assigned projects, but it's probably not a safe call for people who don't have years of experience on a given wiki or who are being asked things that aren't part of their jobs. It's not easy to divide the two, because (for example) volunteer scripts can cause WMF software to look broken. Whatamidoing (WMF) (talk) 02:16, 13 September 2014 (UTC)Reply

Hi Whatamidoing (WMF), firstly it's great to be communicating with you in your WMF-capacity, in addition to on WPMED. What I understand of the current situation is that some WMF employees as part of their employment are expected to engage with the community, and visa-versa the community also expects this to occur. This occurs mainly at certain 'nodes' like this talk page and other talk pages (for example, WT:FLOW). On such nodes (I presume) that one or more members of different teams are responsible for interacting with users as they present problems. For example, the community liason staff, product manager, or nominated team member. As a user, there are times when I would like to interact with the WMF -- eg to provide product feedback, response to a request for comment from the WMF, or respond to a bug. Currently, it's quite unclear where to report that. Providing this small message lets users know that, as part of their responsibilities, a WMF employee is responsible for monitoring that page and responding to queries. If that person is away or goes on holiday etc., then the template can be updated. I see this as just clearly stating something that's already taking place, and would love to hear from you if this is not the case.

With regard to the specifics (such as the 48 hours) I just provide these to illustrate the underlying point about creating a document that clearly states what the community can expect when interacting with staff. I think that one of the big causes of the underlying conflict is a lack of clearly marked venues to collaborate with staff in a timely basis, and a confusion by communication staff as to their role in responding to requests and referring inquiries. For some examples of confusion, who should respond to a query? any member of a team or the community liason? Is the liason expected to 'defend' decisions that have already been made? how vigorously should they engage? (multiple times daily, daily, weekly?) I think the lack of clearly defined guidelines means this is decided individually by each liason, which creates a fair amount of stress. Having some guidelines may help in this respect.

The time limit is a little arbitrary, but whoever is responsible for monitoring and responding probably does need some time limit to respond to inquiries and, if they're not able to respond, at least make a reasonable effort to contact someone who can. I feel, if it is indeed the case that employees are expected to monitor some pages, these are some reasonable responsibilities that are probably already stated in the job or role description. --LT910001 (talk) 09:39, 13 September 2014 (UTC)Reply

Thanks for the reply, LT. It sounds like you're thinking of the fairly minimal (and practical) case: we would tag pages like w:en:Wikipedia:VisualEditor/Feedback, whose main purpose is community–WMF contact, but not pages like w:en:Wikipedia talk:VisualEditor, which happens to be on my watchlist for job-related reasons, but which I may not be paying much attention to, or w:en:Wikipedia:Village Pump (technical), which might be read by a dozen WMF staff on any given day (including me), but whose primary purpose is something other than a way to contact the WMF and it's not officially anyone's job to read it (although reading it might sometimes be necessary for some of the devs' or ops' jobs).
I could see that being useful at larger wikis, where frequent communication is possible. Among other things, finding out where the template is transcluded would give those "in the know" (about the template's existence) a quick catalog of pages to seek staff on. On smaller wikis, it's probably not so useful. I have, in the past, checked in with mid-sized projects a couple of times a week, and small wikis a few times a month, but as there is so little happening at some of them, it would actually be wasteful to check them frequently. In fact, the most useful thing I could have done for some of them would have been leaving a note saying, "If you need help, go to". The urgency and priority for different projects are both variable, so the time limit could be either unspecified or configurable. The time limit could also be framed in terms of what you can expect instead of as a contractual service level agreement: "Most of the time, you can expect a response within ____."
What do other people think? Whatamidoing (WMF) (talk) 23:46, 13 September 2014 (UTC)Reply

Following up[edit]

Hi. I've been absorbing this - there's a lot of information to go through. I do appreciate everyone's patience and flow of ideas. It’s Quarterly review time and while I’ve had my eyes on this page I haven’t had much time to comprehensively respond to the ideas contained.

As Theo10011 and -<(kmk)>- both noted in the first post, this page is hard to navigate. I'm sorry about that - we wanted a way to allow users to brainstorm freely; and I now see that it would have been clearer for all if people signed their ideas. This is an open-ended contribution of ideas, so please feel free to continue to do so.

We’re revisiting the product development process in Quarter 2 (Slide 5) and will be reviewing your ideas continually to see how and where we can integrate collaboration within the product process. As that happens I'd like to work together with you and others within the communities to see which ideas would work most effectively. written, but not signed, by User:Rdicerb (WMF)

And...I forgot to sign the first time. Thanks for adding! :) I've also added a link to the slides from the most recent Metrics and Activities meeting--Rdicerb (WMF) (talk) 22:06, 4 October 2014 (UTC)Reply

Cleaning up some errors[edit]

I've re-organized and corrected a good deal of the first main section on the page. I hope that this will be largely self-explanatory, but if you've got questions, please feel free to ask. In particular, some of the contributors seem to be assuming that Media Viewer is a typical deployment, which is far from true. (I suggest searching the English Wikipedia for pages containing the words "guinea pig" to learn a bit more.)

Also, there's an error throughout this page in which "listen" and "obey" are conflated. The WMF has some (although IMO not enough) mechanisms to determine whether the local community of experienced editors wants a product. (It does less well with finding out what occasional editors and readers want.) Knowing the views of the established editors does not necessarily change the WMF's decision (although in many cases, it does; see most of the resolved bugs in this list for dozens of examples).

I believe it's important for us to be clear about this, because otherwise we'll focus on adding methods of gathering more information from the same people whose opinions are already well understood, and then we'll be right back in the same unhappy place: the WMF knows what The Community™ wants, and the WMF still does not obey The Community™. People who are unhappy about Media Viewer, for example, are not unhappy because of a lack of information being given to the WMF. They're unhappy because the WMF, despite knowing full well that the local consensus (among highly experienced editors) is opposed to Media Viewer, still refused to implement the request from the local community. Whatamidoing (WMF) (talk) 04:00, 12 October 2014 (UTC)Reply

@Whatamidoing (WMF):There was a considerably amount of whitewash included in your editing, to put the bad decisions and behaviour, that definitely were done by WMF, out of focus. Especially the absolute antisocial behaviour towards clear community decisions in deWP and enWP regarding the premature default of a buggy MV got swept under the carpet, and that was a prime example of a complete fuck-up done by WMF. Please consider your POV-editing yourself and put the not-so-pleasant stuff back in there. --♫ Sänger - Talk - superputsch must go 05:10, 12 October 2014 (UTC)Reply
Hi ♫ Sänger,
Pings don't work like that, unfortunately.
I don't think that what I wrote was "blatantly POV". I think the least neutral of your recent changes was this: you declare that people are "rightfully" feeling certain emotions. The fact is that some people feel this; whether it's right is an opinion or value judgment. Someone coming from a different, more authoritarian culture might decide that these people were very wrongly feeling this. I think it would be appropriate for you to remove the word rightfully.
You've also re-introduced falsehoods. The WMF has the ability to stop a project or change the direction of the product. The WMF did not choose to do that for Media Viewer, but it has the ability to do so, and indeed has done so in other cases. For example:
  • The direction of Flagged Revisions was so fundamentally changed in respose to feedback from en.wp, that it ended being a second product. You are familiar with FlaggedRevisions at de.wp; at en.wp, the semi-equivalent product is PendingChanges.
  • AFT (Article Feedback Tool) was substantially revised in response to community feedback several times, and then it was removed entirely, solely because of community feedback.
  • Square bounding boxes was depolyed and completely reversed in the space of just a couple of weeks earlier this year. The reversal was 100% based on community feedback.
Clearly, they do have the ability to hear feedback and respond to it by removing unwanted products: if they didn't, then none of those would have happened.
Similarly, feedback about removal is not "filtered out", although it may be discounted or rejected if it's not responsive to the particular question at hand. The Media Viewer consultation is a good example of that. The question asked was how to make the product work better for people who were actualy using it. "We voted not to let people use it" is not an answer to that question, is it? But even though such a message was non-responsive, that feedback was read and considered by most of the people making the decision.
BTW, certain types of feedback seem to be more likely to result in a product being withdrawn. If you'd like to talk about that, then let's make that a separate discussion sometime. Whatamidoing (WMF) (talk) 16:14, 12 October 2014 (UTC)Reply
A more authoritarian culture is out of question, that's bad to the bone. Anyone even contemplating such things is not fit for the wikiverse, there is no discussion necessary about that. Superprotect, or better named superputsch in the first use case, was authoritarianism pure, and thus bad to the bone, anyone involved in that action against the community has to be reprimanded because of this. There was no justification for superputsch besides the hurt ego of the bling-pushers.
The WMF has the possibilities to do an lot of stuff, just like any junta general has the possibilities to use a lot of weapons. It just has no right to use any tool for any reason, especially not against the real sovereign, that's the community, or in case of the junta general the elected government. Erik (and as accessories the whole WMF that failed to reign him in) behaved like any junta general, just far less bloody as this is just a virtual world, but of course he expulsed a lot of good editors and admins with his repugnant acts.
For MV the question was: Is it fit for opt-out or not? And the community decided: Not yet. But the WMF didn't want to hear that answer, so it just didn't listen and shoved it with might against the communities. I'm still amazed why they did so much harm because of this so unimportant bling-thing. They even admitted there are some "must-have" things not yet implemented, and still refused to make it opt-in again, imho a no-brainer. Nothing with ans "must-have" features still not working must ever be let loose on the whole community, the minimum condition has to be: All "must-haves" are in there, otherwise it's still just for testers.
Fabrice contradicted himself in one paragraph, and I think he didn't even noticed in his warped mindset: "The consultation outcome was not at all pre-determined [...] And we were very clear from the outset that requests to turn off the software were outside the scope of this consultation." It can't be both, that's newspeak at its best.
The WMF wanted to restrict the discussion to some superficial nonsense, and the same arrogant mindset seems to surface again with Flow, some up to now useless blabber tool, that's massively promoted as the next "core feature", while it's not fit for anything yet. If the WMF behaves further so active against the communities that deliver the donations with their content and maintenance of the wikiverse, and thus ultimately pay all the WMF-staffers, I don't know who will further want to work in such an environmment. In that case the not-fit-for-collaboration WMF should fork without the donations and found their own environment. --♫ Sänger - Talk - superputsch must go 17:23, 12 October 2014 (UTC)Reply
We seem to be going a bit further afield, but let me reply to two of your points:
  • A "more authoritarian culture" is exactly what some of our users and some of our communities live with every day. I don't want it, you don't want it, but that doesn't mean that no editors from those cultures exist, or that editors who do live in (and even support) those cultural values would not have a more accepting reaction than us to these events. I believe that editors from those cultures are fit for the wikiverse; I even believe that the world would ultimately benefit from having far more editors from authoritarian cultures like North Korea, Saudi Arabia, China, Cuba, and Singapore (because the way to get those editors is to have far, far more unfettered access to Wikipedia in those cultures).
  • The outcome of the Media Viewer consultation was not pre-determined: if truly severe technical or legal flaws in the software had been identified, they were going to pull it from all the wikis, including the projects that wanted it. However, no truly severe flaws in the software were found. "We voted to remove it" is not a technical or legal flaw. As the consultation page said repeatedly, the question being asked was not whether you wanted people to use it, but whether there were problems in the actual software itself that would harm the people who were actually using it. Whatamidoing (WMF) (talk) 18:53, 13 October 2014 (UTC)Reply
Ugh, so you now suggest to make decisions around here based on judgements about populations? I'd love a source for «authoritarian cultures like North Korea, Saudi Arabia, China, Cuba, and Singapore» (emphasis added). --Nemo 15:23, 14 October 2014 (UTC)Reply
No: I suggest that different users will have different reactions, and that frequently individual's reactions will approximately line up with their own culture. I also state that not every user will have Sänger's reaction (or mine).
(As it happens, I checked every one of those examples. Putting "authoritarian culture" <name of country> into your favorite search engine will find you plenty of reliable sources.) Whatamidoing (WMF) (talk) 17:00, 14 October 2014 (UTC)Reply
I couldn't find a single source on Google Scholar for your proposed keywords for Korea. I invite you to absolutely refrain from attaching labels on people based on their geographical origin, as long as you're commenting on a Wikimedia wiki. --Nemo 06:21, 22 October 2014 (UTC)Reply
Try Google Books, and please notice that I specified "North Korea", not historical or traditional Korean culture. Whatamidoing (WMF) (talk) 16:35, 22 October 2014 (UTC)Reply
I haven't really dug into this page and the subject-space page, but edits such as this most certainly do have the appearance of whitewashing, particularly given that the edits are coming from an official Wikimedia Foundation account. I have a lot of difficulty looking at edits such as this and finding them appropriate.
Honest and open criticism about feature prioritization and resource allocation is highly important and we should be doing our best to recognize and address these very real issues rather than simply pointing to the fact that the criticisms aren't necessarily universal. In other words, whether or not MediaViewer is/was atypical is irrelevant: the reality is that certain projects/products proceed without a clear (or any) request from any Wikimedia community and killing or changing these projects/products can be nearly impossible. And there's nothing wrong with saying so. --MZMcBride (talk) 19:41, 12 October 2014 (UTC)Reply
Max, I'm not sure that an honest and open conversation is possible unless we can agree on some basic facts, like "It is a fact that the WMF has actually reversed deployments in response to community requests (although not this last time)" or "It is a fact that most new products are usually not deployed to small wikis first (unlike this last time)". You know far more about the deployment history than most people who have been putting forward their ideas so far, and I know that you'd like to have an accurate description of reality.
I agree with you: some projects proceed without requests from any communities, and some products are continued despite requests from one or more communities to stop. There's nothing wrong with saying that. There is, however, something wrong with saying or implying that all products (or even most products) have these characteristics. In fact, I believe that you would agree with me that, considering the history of the last few years, it's fairly rare for a community-suported configuration change to be refused—even though that's exactly what happened with Media Viewer. Whether Media Viewer is atypical does matter: we are unlikely to design a system that works for most cases by claiming that software deployments almost always happen like Media Viewer did. Whatamidoing (WMF) (talk) 18:53, 13 October 2014 (UTC)Reply
Quite true, it's important to get facts straight first. For instance, "It is a fact that most new products are usually not deployed to small wikis first" is false. I suggest that you familiarise yourself with the deployment process; I believe you might be able to do so starting at wikitech:Deployments and mw:MediaWiki 1.25/Roadmap. --Nemo 15:29, 14 October 2014 (UTC)Reply
Nemo, I'm a bit confused by this comment. Do you believe that:
  1. new products are normally deployed to (only) small wikis first, or
  2. new products are normally not deployed to (only) small wikis first? Whatamidoing (WMF) (talk) 16:45, 14 October 2014 (UTC)Reply
Neither, or (1), depending on definitions. See subject page. --Nemo 06:21, 22 October 2014 (UTC)Reply
Okay, you believe that new products are normally deployed only to small wikis first (assuming that it's not everyone at once and naturally excluding all test wikis, including
I don't agree with that assertion, but if you believe that, then can you tell me why you write on the project page that "many other products (e.g. AFT, MoodBar, PageTriage, Echo, Flow) are deployed on the English Wikipedia as first content project, as happens for experiments"? "It's normally the small wikis first" is not compatible with "many products reach the largest wiki in the world first". Also, if you believe that the process is normally to deploy to small wikis first, then what is the point behind proposing that the WMF stop deploying first to en.wp? According to your response here, the WMF is already deploying software to places other than en.wp first. Whatamidoing (WMF) (talk) 23:01, 24 October 2014 (UTC)Reply

Never first[edit]

(I'm copy pasting here some comments about this matter from the main page. --Elitre (WMF) (talk) 11:59, 21 October 2014 (UTC))Reply

Hi there - as far as I'm aware, English Wikipedia is traditionally one of the last projects that the WMF releases to - at least during this point in time. For example, the Media Viewer release plan outlined that English was one of the last deployments. I would love to documentation of projects first deployed to English Wikipedia so that I can better understand the issue. Feel free to leave a message on my Talk page or email me with information/links to relevant release information. Thank you! --Rdicerb (WMF) (talk) 23:14, 14 October 2014 (UTC)Reply
I covered this in #Current state: «"Big" feature development (e.g. AFT, Echo) and experiments are often deployed on the English Wikipedia first». How many examples (3, 10, 50?) spanning how many years (1, 2, 4?) would you need? --Nemo 18:01, 15 October 2014 (UTC)Reply
I believe (but please don't quote me on that!) that there were several cases in which en.wp wasn't the first wiki: LiquidThreads, FlaggedRevisions, Page Curation (possibly more). The very fact that the latest product (MV) follows this pattern is IMHO quite interesting/revealing, but this said, it really depends on the products. Some communities have more pressing needs so they may sometimes come first. (I am not a fan of "set in stone" rules, especially when it comes to products. I don't think it works well in our world.) --Elitre (WMF) (talk) 18:02, 16 October 2014 (UTC)Reply
The English Wikipedia has been the first production site for deployments for years. The English Wikipedia is convenient for devs (all of them speak English), it's big enough to get a lot of testing down quickly, it's easy to get well-informed feedback from the tech-savvy editors, it's the home wiki for many of our volunteer and staff devs, you can find out what happens with complex environments (en.wp articles are the most complicated and varied), and it's the traditional site for deployments (partly because, once upon a time, it was the only wiki, and the habit carried over).
It would actually be faster to make a list of what wasn't deployed at en.wp first before the CLs were hired. Erica's got most of the significant items in her list in the previous comment. I'm not certain that Page Curation, which was designed for the English Wikipedia, belongs in that list. LiquidThreads was never used on the English Wikipedia because it couldn't scale to that size. FlaggedRevisions wasn't deployed because of community objections. It was eventually re-written as PendingChanges. Whatamidoing (WMF) (talk) 21:31, 21 October 2014 (UTC)Reply
Every single thing you wrote here is factually incorrect (with exceptions in the "The English Wikipedia is..." sentence) and shows an extremely limited view of our software development process. Luckily, I've improved the subject page now, which should facilitate learning. --Nemo 06:21, 22 October 2014 (UTC)Reply
I'm not sure who you are replying to. As there are almost opposite statements above, it's unlikely that it applies to all of the above. Since what I wrote agrees with what you wrote (that [whenever it's not all Wikipedias at once] it's most frequently en.wp first), it's unlikely to be a reply to me. Whatamidoing (WMF) (talk) 16:31, 22 October 2014 (UTC)Reply

Product priorities survey idea[edit]

As the WMF continues planning priorities for upcoming quarters, we are looking for more ways of engaging communities in broadly influencing the product roadmap. At this point there have been community initiatives to list requests, such as with dewiki’s Top 20 list of requests. In order to ask what you want us to build across the movement, I’m looking for a way to use and quantify a centralized, unified list that anyone from any part of the movement can contribute to, so that ideas can be captured and ranked. It will help the Foundation to know what your primary product concerns and focuses are.

Current on-wiki systems like Talk pages are not as accessible to less experienced users (who may face some of the same issues as power users, such as edit conflicts). Discussing in one location like Meta may not bring in the broad range of users that we’re hoping to reach. Also, to ensure that the ideas that you think are the most important get attention, a measurable system is needed to accurately rank and weight your input and feedback. We’ve been looking into a few survey systems in the past couple of weeks to try to find one that might work with the kinds of criteria that are needed (good privacy policy that anonymizes your input, allows you to add ideas, preferably open source).

The goal is to decide on a survey system in the next couple of weeks, then test with interested users to ensure that the system cannot be easily gamed and that it works. Then we would consult on a broad scale in Q3.

What this would be:

  • an attempt to bring more community influence into the Foundation’s product roadmap by ranking submitted ideas
  • a means to encourage a broader range of users into telling us what their needs are
  • a means of including data directly in the process

What this would not be:

  • a replacement for any products currently in development
  • a promise that all ideas submitted will be built

I'll update this thread with more details as things move further along. Let me know if you have any questions. --Rdicerb (WMF) (talk) 18:52, 28 October 2014 (UTC)Reply

  • Is multilingualism not one of the requirements?
  • As usual, it would be nice to avoid duplication. Suggestions are potentially thousands: it's expensive for users to share them and for [whoever] to collate them. Bugzilla is imperfect but is where most of our treasure of software/product ideas is. So, for instance, you could allow free text ideas, but also prefill with some dozens or hundreds or thousands options from bugzilla; carefully selected perhaps from the reports with more than N votes (again, votes mean little but are an efficient tool to focus when you want to manually examine only 10, 100, 1000 reports rather than 50,000).
  • Erik in 2010 suggested IdeaTorrent, is that the sort of thing you have in mind? We've also had recurring discussions on (Q&A site) (and how to turn bugzilla & co. into one such thing). --Nemo 12:31, 5 November 2014 (UTC)Reply
I fully agree with Nemos comment about multilingualism and duplication and the expense for editors.
  • "We’ve been looking into a few survey systems in the past couple of weeks (...) The goal is to decide on a survey system in the next couple of weeks" So, where is the list of these "candidate" survey systems? --Atlasowa (talk) 16:29, 5 November 2014 (UTC)Reply
I would also like to see a candidate list. — Jeblad 19:23, 5 November 2014 (UTC)Reply

Thank you both for the nudge and the questions. For language support: Yes. Absolutely. We are running a small scale pilot in Q2 with 2 languages just to try out to see if the system works, then launching on a full scale in Q3. Languages would be English, and I was thinking either Spanish or French. Thoughts?

For this first version, we'll be asking about expanding a gadget or tool into an extension.

For the criteria for a system, I was looking at the following:

  • Language support
  • Does it support segmentation (by tenure, edit count, etc)
  • Is it open source?
  • Language (PHP, Python?)
  • Must allow contribution
  • Does it rank/tally ideas?
  • Must hold many ideas
  • Robust against gaming
  • Proven scaling and support (who fixes it when it goes down?)
  • Compliant with our privacy policy

For the candidates, it was narrowed down to a group at Stanford going by Synapp (this system would need customization to fit our needs) and All Our Ideas, which was used by the WMF in 2010. Nemo, I'd like to launch this by Dec 1st for a 2 week pilot so that we can do a broader launch in January, but you mentioned Idea Torrent which we have not looked into. I'll check it out. Page for this project coming soon (looking at my schedule...hopefully tomorrow afternoon?).

As for the sentiment of recreating work: while I do not have the institutional knowledge that you all have, I do see there is sometimes the tendency to recreate work and I would very much like to not do this :) We're looking at expanding a gadget or tool, so I will clarify that in the (coming soon!) project page. Perhaps we can gather a list of lists on that discussion page when I post it? -Rdicerb (WMF) (talk) 02:19, 6 November 2014 (UTC)Reply

Late/brief update: Some of this work is still ongoing at phab:T94807 ("Identify survey services compatible with our privacy policy"). But depending on the individual use-case, will obviously need to be further measured against the criteria Rdicerb listed above, and that others had discussed in the other linked locations. There was also further discussion at Talk:Community Engagement (Product)/Product Surveys in December 2014. Quiddity (WMF) (talk) 20:57, 4 August 2015 (UTC)Reply

JS gadgets, Lua scripts, templates[edit]

I would really like to see central repo for Javascript gadgets, Lua scripts and possibly internationalized templates. All three would be crucially important for mid-sized and small languages, as they most likely lack people that can adapt and implement solutions in those areas. As we slowly are starting to loose manpower we must be more efficient. It is not a viable solution to keep on insisting that every project do all the kind of stuff themselves. We must start to consolidate and reuse work wherever possible. The listed three areas are all areas where we can ruse work done on central high-volume sites. It is although important to have some services available to make such a repo work, like a lint service and a test service. It should in other words be possible to develop something, check it with lint and test it, and then if it still passes can it be allowed to go into production. We must have something like FlaggedRevs for gadgets, scripts and templates, that is we only set stable versions in production. — Jeblad 19:37, 5 November 2014 (UTC)Reply

@Jeblad: There is some discussion of the various possibilities around ideas like this, at phab:T71445 and links within. There was also a recent post at that might interest you, if you hadn't seen it, that touched (in part) on a few aspects of templates. HTH. Quiddity (WMF) (talk) 21:14, 4 August 2015 (UTC)Reply

Moved comment[edit]

I have removed the following comment from the content page, since it belongs here instead. Neil P. Quinn (talk) 04:41, 6 February 2015 (UTC)Reply

„... there seems to be no ability to accept input that a project should be halted or fundamentally change direction.“ Please compare the understanding of „vertical tolerance“ during the 3th till 14th century. --Edward Steintain (talk) 19:00, 18 December 2014 (UTC) You are right as long as you agree – otherwise your „soul has to be saved!“


I'm desperately trying to find a way to return opensearch results from our wiki in this exact format:

<rss version="2.0" xmlns:opensearch="" xmlns:atom="">
<title>VHD Search</title>
<description>Search results for at VHD</description>
<atom:link rel="search" type="application/opensearchdescription+xml" href="">
<opensearch:Query role="request" searchTerms="" startPage="1"></opensearch:Query>
<description><br>The Goldsborough Mort Building was built in 1862 to designs by architect John Gill and was Richard Goldsborough's first warehouse in what became an extensive complex in this area of the city...</description>
<description><p>The Cobblers Gully Gold Puddling Site consists of the remnants of one puddling machine complete with an adjoining water dam...</description>

Unfortunately, after installing the opensearch extension onto our wiki, we can still only return results in this format:

<?xml version="1.0"?>
<SearchSuggestion version="2.0" xmlns="">
  <Query xml:space="preserve">irish</Query>
      <Text xml:space="preserve">Irish Famine Orphan Immigration</Text>
      <Description xml:space="preserve">The “Earl Grey” Scheme was implemented to select female orphans from the workhouses of Great Britain and Ireland and to bring them to the colonies. </Description>
      <Url xml:space="preserve"></Url>

We are running semantic mediawiki v1.19. Is there anything we can do to solve this problem?

Thanks, Asa

Hey Asa, there are 3 ways to get help about that:
Hope this helps. --Elitre (WMF) (talk) 08:57, 10 March 2015 (UTC)Reply


Has any progress been made with this project? Or has it just died? Some feedback on what if anything has happened or is likely to happen would be welcome. 11:16, 6 April 2015 (UTC)Reply

FYI: Special:Permalink/11775298#Process ideas update? -Pete F (talk) 03:27, 12 April 2015 (UTC)Reply
Hi there - when this page was created it was done so in a way that lacked a structure, and workflow priorities have gotten in the way of making this fully implemented (as Pete F notes above). It's possible that we may restructure this page (there were comments early on that it seemed to lack a known structure). The primary question here has been around integrating community feedback specifically into product and engineering development. For example, a couple of new mechanisms for giving feedback right now on VisualEditor are the Editing Team's public bug triage meetings - there was low community turnout in Q3, but they're held at a fixed time, and anyone is welcome to attend. Community Liaisons are also asking those who are either frequent users of VisualEditor or who have given feedback to the Foundation in the past to answer a brief survey about it (closing soon, so if you have not participated and would like to, please do so before Thursday afternoon Pacific Time) - essentially, engaging users in product development and continuously seeking ways of gathering feedback.
I think there are certainly some good ideas in this list, and I apologize that a whirlwind of tasks hasn't allowed for feedback on this, and I think there are things we're already doing. What concerns me (and of course you're welcome to disagree) is that adding to this page might make it harder to follow at this point. That being said, going through and confirming what can be done, what is being done, and what cannot be done on each point would be helpful, I agree, so let me try to get this into my list by the end of next week (if anything I will give a timeframe by then). -Rdicerb (WMF) (talk) 02:45, 14 April 2015 (UTC)Reply
I didn't make a point above -- all I did was to link a related point on a user talk page. I don't have any special insight into the shifting priorities at WMF. -Pete F (talk) 16:54, 14 April 2015 (UTC)Reply

Check in Late April 2015[edit]

Tracked in Phabricator:
Task T97187 open

It's true, we have not been checking in as much as we should. We're trying to change that (pings to my talk page are always welcomed). The team is going through this list and updating on the content page itself. There's a Phabricator ticket about this; you're welcome to follow and discuss there as well. -Rdicerb (WMF) (talk) 23:52, 27 April 2015 (UTC)Reply

Change discussion venue, do not respond directly, change the topic... - Amgine/meta wikt wnews blog wmf-blog goog news 02:44, 28 April 2015 (UTC)Reply
Since this section was started in April, nothing has been added to the list by any team member and no progress has been reported at the Phabricator page. The statement "The team is going through this list and updating on the content page itself" is false. It seems clear that this project has failed and is defunct. @Rdicerb (WMF): if you wish to persuade the community otherwise, it will be necessary for your team to actually take some action. The time for promises and excuses is over. Rogol Domedonfors (talk) 11:59, 6 June 2015 (UTC)Reply
The absence of any response, or even acknowledgement, on a page being actively monitored by a responsible member of WMF staff, is not merely unsatisfactory, it is unacceptable. @Rdicerb (WMF) and Whatamidoing (WMF): I regard the failure of you and your team to act or explain your inaction as an indication of a level of contempt for the community sufficiently grave that I regard it as misconduct. I do not wish to be unjust and so wish to allow a brief window of opportunity for you to persuade me that I should not ask for management to take disciplinary action in this case on the grounds of neglect of or incompetence at your jobs. Rogol Domedonfors (talk) 12:07, 21 June 2015 (UTC)Reply
My assignments are keeping me busy with other things, and finishing the responses for this page has never been one of my assignments.
The template at the top of the page was added by a volunteer editors, and I'm afraid that it's a bit misleading. It was meant to be a test of a concept. I think I can now report that it works, which is nice, but the circumstances for proving that are less than optimal.  :-/ I'll remove it. Whatamidoing (WMF) (talk) 23:35, 21 June 2015 (UTC)Reply
@Whatamidoing (WMF): you have read and contributed to this page on numerous occasions during the nine months that it carries a prominent notice claiming that you were active on it. Now that a volunteer criticises your lack of action you disclaim responsibility and claim that the notice, of which you were well aware, was misleading. This in connection with a project that you were very active in, while you held, and continue to hold, a community liaison post. No doubt senior management will give your explanation the weight it deserves. Rogol Domedonfors (talk) 06:29, 22 June 2015 (UTC)Reply
If you take the position that community is the employer of the Foundation's employees, then your behaviour here constitutes employee harassment. Feel free to take it up with the superiors, but leave the workers to work. - Amgine/meta wikt wnews blog wmf-blog goog news 15:55, 22 June 2015 (UTC)Reply
... and stopped again the same day with two out of 35 suggestions annotated, a rate of progress which if maintained would suggest that the analysis will be complete in almost exactly a year's time. Is there a planned finishing date for the completion? Is it earlier than August 2016? Do you propose a more sophisticated analysis? Rogol Domedonfors (talk) 16:19, 23 August 2015 (UTC)Reply
Can we just declare this project dead now? The discussion has moved away, there have been major changes to the WMF internal organisation, a new strategic plan, a new community tech team, changes to major projects, ... the chances are low that this is anything other than a set of ideas to refer to for inspiration during the odd brainstorming session. The correct application of staff and volunteer time and effort is to determine why it failed and how to do better in future. Rogol Domedonfors (talk) 21:35, 6 September 2015 (UTC)Reply


I have marked this page as {{historical}}. Repeated requests, dating back several months, for an update on its status, or a review of its contents, have gone unanswered, and any further contributions by members of the community seem unlikely to have any practical value. Rogol Domedonfors (talk) 18:28, 10 July 2015 (UTC)Reply

Reverted, per what I previously said. Said "imposed" tagging isn't respectful; nor truthful, as the last update was less than three months ago.
Please stop it, you've been the only one advocating for this edit in multiple discussions you opened in multiple talk pages (in a rather annoying cross-posting). --Nemo 19:46, 10 July 2015 (UTC)Reply
If you're going to use the word "truthful" you need to pay careful attention. The diff you give is an update to this page, but is not "an update on its status" which is what I said. However, you are perhaps right, and I appreciate the corrective, about whether it is worth pursuing. If a poorly planned and managed project achieves rather little and is left to wind down unattended without resolution or output -- why should I devote my time and energy to fixing it? Rogol Domedonfors (talk) 20:43, 10 July 2015 (UTC)Reply
It turns out the project has not been completely abandoned: see the next section. I am happy to acknowledge that. Rogol Domedonfors (talk) 15:50, 25 July 2015 (UTC)Reply

August 2015 Engineering management offsite[edit]

On the subject of this project: I was told by the Executive Director today [15] that We are working on a draft that we intend to publish this quarter on improvements to our software process. Engineering management has an offsite specifically focused on that at the end of August. I am grateful to her for that information, which I thought would be of interest to anyone still reading these pages.

Please would a staff member update this project page with the date of the offsite, or, better, the last date by which community comments will be able to be considered at that meeting. It would also of course be much appreciated if when that draft is published, a link to it could be put on this project page, together with a link to the page where you will host community discussions. Rogol Domedonfors (talk) 15:50, 25 July 2015 (UTC)Reply

According to phab:E46 it is 26--27 August. Rogol Domedonfors (talk) 16:24, 23 August 2015 (UTC)Reply

Community Engagement (Product)/Wikimania 2015[edit]

Relevant discussions continue here. Rogol Domedonfors (talk) 20:22, 7 August 2015 (UTC)Reply


November 8, 2015 Superprotect functionality was removed from Wikimedia servers.[16]

A much appreciated step to improve WMF-Community relations. WMF staff may wish to edit or remove the superprotect&de-sysop threats that still linger in section: Requested changes? Alsee (talk) 06:47, 8 November 2015 (UTC)Reply

Pinging WMF staffmember with the most edits on this page: User:Rdicerb (WMF) Alsee (talk) 10:45, 26 December 2015 (UTC)Reply

Done Thanx. Alsee (talk) 13:54, 1 January 2016 (UTC)Reply