Talk:Wikimedia Foundation Medium-term plan 2019

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search
The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section.

the addition of new form factors[edit]

What does "the addition of new form factors" mean, and is there another way of saying this that works equally in American English and English? I'm guessing it is an American English phrase. WereSpielChequers (talk) 12:45, 6 April 2019 (UTC)

Hi WereSpielChequers -- The definition of form factors we're using is from the perspectives that the Audiences team has written "The term Form Factor generally refers to the various entry points, devices, channels and formats that define a digital product offering. In the context of Wikimedia, form factor will include (a) the variety of devices that Wikimedia content can show up on now and in the future; (b) the size, and flexibility of the content itself. Such explicit form factor considerations are the main focus of this paper, however, other implicit aspects of form factor must be considered as well. Partnerships, for example, may require Wikimedia content to be adapted to, or deeply integrated into, third party products with assistance and guidance from the makers of those products - and several of the explicit considerations suggest this type of partnership. Likewise, ideas like making Wikimedia content available as a utility, or layer on top of the Internet, are another implicit example of form factor explored in some of the concepts described in the paper."
I'd definitely recommend reading the perspective linked above if you have more interest in this! TNegrin (WMF) (talk) 18:54, 10 April 2019 (UTC)
Or the WMF could ask for a plan that is written in plain English and where necessary defines more complex and specific words in an accessible way. Nobody can expect unpaid volunteers to spend a week trying to mug up on a library of background documents before being able to actually "read" the top level plan. -- (talk) 12:27, 13 April 2019 (UTC)
Others on this page have pointed out the need to avoid jargon in a document like this for an organisation like us with an objective like ours. Translating "form factors" into "whatever devices people use to access the internet, we need to make our sites compatible with their devices" should hopefully be uncontentious. The contentious area, the one that is likely to trigger the communities CC0 CC-BY_SA divide is over deeply integrated and adapted uses of data. In theory this shouldn't be an issue, our content is not licensed ND. But in practice the people pushing for adapting and deep integration of our contributions are usually the people who struggle with either attribution or share alike. I wouldn't be so concerned about this if the Foundation were to adopt enforcing BY and SA on behalf of the community as a main strategy of motivating editors, recruiting new editors and keeping access to donors. But that doesn't seem to be there, in fact it looks like the biggest gap in the strategy. WereSpielChequers (talk) 17:09, 16 April 2019 (UTC)
Form factors seems to also be dependent on the reader used by the mobile phone to fit the article onto the small screen. Per a recent report, in the Global South, mobile viewing is the principal method. Geraldshields11 (talk) 13:26, 18 April 2019 (UTC)

ensure content is locally relevant to communities.[edit]

Our projects are almost all language based, some of those languages are local, some national some much more so, and a few are global. I get the idea that the Aceh, Navajo and Georgian wikipedia's are primarily going to be of interest to particular communities of at most a few million people in one area. whilst English, Spanish and Arabic are covering very substantial parts of the world. Is "ensure content is locally relevant to communities" a dig at the likes of Cebuano, trying to create a complete copy of Wikipedia for a medium sized language group? Or is it an intent to balkanise large Wikipedias and create national or sub national ones for languages such as Arabic and English? WereSpielChequers (talk) 13:08, 6 April 2019 (UTC)

Hi WereSpielChequers, Thanks for the question. I can see how the term "local" could be confusing. There are a few layers to what we mean here. Let me give some examples. First, we know that English Wikipedia is missing critical content in primarily English speaking countries, such as Nigeria. We want more content that is relevant to Nigerians in English Wikipedia. Also, we also know that the encyclopedia in languages such as Hindi (with over 400 million speakers) is not nearly robust enough. We want to grow content in these languages and include knowledge that is specific to that culture as well as global knowledge. "Local" is probably not the best shorthand term for this, so please share if you have any ideas for a better way to describe it. Thank you. --Lgruwell-WMF (talk) 21:03, 10 April 2019 (UTC)
Hi Lgruwell-WMF and thanks for the explanation. If this is about filling in key gaps rather than curating/censoring for particular "sensibilities" then I'm with you. Pretty much all of our content is irrelevant to somebody, most is irrelevant to almost everyone. But if we rephrase this to filling in gaps without threatening some huge introspective curation exercise then we avoid the whole inclusionist/deletionist debate. I suggest "get our readers the content they most need in the language they want it in, remembering there are large parts of the world currently coming online". WereSpielChequers (talk) 09:53, 16 April 2019 (UTC)
This goes back to Worldwide Readership and Thriving Community. If WMF could promote a safe space thriving community to add more local content that meets Wikipedia standards, then the readership, in their own language, can benefit. It is a cost/benefit analysis to recruit volunteers to fill "local" needs. The goal is noble and necessary but I think it needs to be stressed more in the Mid-term plan. Geraldshields11 (talk) 13:32, 18 April 2019 (UTC)

competitive in an ecosystem in which machines creating content,[edit]

I suspect "competitive in an ecosystem in which machines creating content," should be "competitive in an ecosystem in which machines are creating some of the content,". WereSpielChequers (talk) 13:12, 6 April 2019 (UTC)

Yes, agreed. --Lgruwell-WMF (talk) 21:07, 10 April 2019 (UTC)
On LinkedIn, the Wikimedia Foundation posted "Can machine learning uncover Wikipedia’s missing “citation needed” tags?" on 3 April 2019. Do you mean that type of AI? Or, do you mean AI will create new articles based on third-party sources? If the later, I have the concern that an early AI will create something similar to the first film written by AI debuted, Sunspring, starring Thomas Middleditch. Geraldshields11 (talk) 13:46, 18 April 2019 (UTC)
We have had bot creation of stubs since the early days of Wikipedia. Rambot's creation of articles on populated places in the US is one of the key events that built the project. Maybe AI has got more advanced in the last decade. WereSpielChequers (talk) 20:08, 18 April 2019 (UTC)

In my view over the plan I see this in my opinion really critical point: We know that at the end of the pipeline ever must be an human, but that could be enought: We know that AIs could get own priorities and deceive their humans. If some information that come to the human sitting at the end pipeline is produce by AIs, it could be that it has many inflouence about what happend at this end of the pipeline. This is also showed by why we use this: It makes for us easier find information, and the creation which control information control the decision. So this need very mush ethical control, because for an AI, also with help, there is no difference between the human as only a number-objects or a real subject. Control options for this are missing in this draw. Habitator terrae (talk) 17:13, 18 April 2019 (UTC)

Growth and the PC/Smartphone divide[edit]

There are some welcome growth targets, but they lack a couple of important dimensions. Where Wikipedia isn't blocked, we have a very high readership among experienced internet users. But our ratio of editors to readers is very different among PC users and smartphone users. wikipedia is close to being a broadcast media for those on Smartphones. The Foundation can only measure but not really alter the proportion of humanity who have internet access. Though it can help via CD get copies of wikipedia to people who have access to computers for people who are offline. It would be helpful for projecting Wikipedia growth to work out how long new Internet users usually take to find Wikipedia, and how long it takes them to start editing wikipedia. We know that editing is not an entry level task on the internet, but we don't yet know how long it typically takes wikipedians from when they become new Internet users to when they start to edit Wikipedia. Presumably that will vary depending on whether they have PC access or Smartphone access. We need to know what those typical lead times are before we can identify which societies are slower at using Wikipedia and therefore require extra intervention. Needless to say, improving the mobile site as an editing interface is one of the most important things that the foundation could do if it wants to recruit more editors. It may even be more important than fixing the edit conflict problem, moving to autosign on talkpages or even removing capcha for newbies adding referenced material. So I would suggest rewriting that part of the strategy, if the Foundation wants to recruit more editors, the easy and effective way to do so would be to fix some of the software barriers we currently have to keep the community managably small. WereSpielChequers (talk) 18:27, 6 April 2019 (UTC)

Hi WereSpielChequers -- this is a complex question, so I'll just answer part of it now. We've made substantial investments in mobile editing -- project pages are here, here, here and here if you want to take a look. We're also looking carefully at the first 24 hours after users register to find improvements in the onboarding process. We're hopeful that these efforts will start showing some impact in this fiscal year and also in the future. I *am* concerned that technical solutions may not be enough -- I think we also need to help new users understand the culture and norms of the editing communities as well as encouraging these communities to be more welcoming to new users. TNegrin (WMF) (talk) 19:16, 10 April 2019 (UTC)
@TNegrin (WMF): On mobile editing: Is it genuinely impactful, content-wise, to improve mobile editing? The vast majority of good content on the English Wikipedia is still written with the same old HTML text box-based editors, so I would agree with WereSpielChequers that it has always seemed to be an odd focus. w:en:User:Cullen328/Smartphone editing might also be insightful – I've made a few hundred edits on mobile, including to Lua modules, but I've always used the desktop editors (2010/WikiEditor and 2011/CodeEditor). The Timeless skin helps because its layout is mobile-friendly (as is the Monobook rewrite's layout).
On the Growth team work: Is only the first day being looked at? While the initiatives may have positive results, it seems to me that this would be somewhat futile, given that most contributors who make a genuine impact (at least on the English Wikipedia) usually make at least several thousand edits over their tenures, and they usually edit for at least a few years. Is it too soon to tell if that project's work is resulting in better long-term editor retention? (I've already made some related comments below. I've probably written too much on this talk page.) Jc86035 (talk) 14:16, 15 April 2019 (UTC)
Hi Jc86035 yes if we want to address our ethnicity skew (probably the biggest demographic problem in our editing community) we need to make mobile editing more viable. we have a huge ethnicity skew in the editing community and this makes it difficult to achieve our objective of making the sum of human knowledge freely available to all. In many countries the smartphone is currently the main or even sole editing device. How we improve mobile editing I leave to programmers ergonomists and mobile phone users, that we should aspire to improve mobile editing is something I am very keen to see at the heart of the strategy. WereSpielChequers (talk) 10:02, 16 April 2019 (UTC)
@WereSpielChequers: I suppose this is the dominant perspective, that of increasing diversity to allow for a greater breadth of content, which I do agree with, although (as I've noted below) it might be appropriate to approach this from an ethical perspective as well: if someone doesn't have access to a computer with a physical keyboard, should they still be encouraged to essentially work for free?
While ethnicity is certainly a factor in the breadth of content, it's also probably important to consider that the articles [and other content] that we can produce are dependent on the diversity of points of view in reliable secondary sources, and that diversity might be less of an effect here than in other fields due to the somewhat more limited creative license afforded by the medium. Furthermore, because of cultural and economic factors, which (as you're no doubt aware) are presently inextricably linked to gender, ethnicity and nationality, those who are relatively privileged almost certainly do have a leg up on how able they are to edit Wikipedia [and the other projects], and the factors would also affect how likely they are to edit Wikipedia, and how much time they can spend editing. Most of these factors are, needless to say, currently out of the WMF's control. (...Also as noted below. Sorry if I'm being repetitive, though I do think it's important to consider these things.) Jc86035 (talk) 10:23, 16 April 2019 (UTC)
Hi Jc86035, if the objective were to recruit the same proportion of volunteers everywhere then I'd agree with you. But making mobile an edit friendly medium is more about removing a barrier and is within the WMF's control. Absolutely there are other barriers, and also secondary motivations, we should never underestimate the value of editing Wikipedia in a language you aren't completely fluent in as a hobby that also improves your language skills. This does give us an obligation in English and possibly other languages to be nice and tolerant to those whose English is not brilliant. It also doesn't help our problem with languages that don't have big learner communities such as some of the south Asian and African languages. But we shouldn't assume that all poor countries lack leisure time and all developed countries have it. WereSpielChequers (talk) 17:33, 16 April 2019 (UTC)
Dear everyone, Yes, removing the barriers to editing is what Grow participation globally is all about. The other barriers/factors Jc86035 mention are there but at least WMF could take a step to help the casual editor to edit on the go. Geraldshields11 (talk) 14:07, 18 April 2019 (UTC)

Increase Readership by approximately 20% over the next 5 years[edit]

Our readership is a substantial subset of the Internet userbase, and that userbase is growing fast, if the past decade is a good indication much much faster than 20% over the next five years. Even in the most mature countries for Internet usage there is still some growth as the last of a pre internet generation shuffles off the mortal coil, but most of the next five year's growth will be among people who don't speak English, a large proportion may not be literate, or at least not in a language that is well covered by Wikipedia. I'm not sure where the 20% figure comes from, but judging from past growth of the internet it isn't anything close to the expected growth of the internet in the next five years. I would suggest replacing it with the more challenging "We aim to increase our readership as fast as the total readership of the internet increases". How that is done is a matter for tactics, some languages may need their wikipedia kickstarted. for other languages we may need a query process and text to speech arrangement that allows mobile users who have never learned to read and write to ask for a wikipedia article and have it machine read to them. other languages such as Georgian just need script support so that more users of that language can input Georgian script via Qwerty keyboards. WereSpielChequers (talk) 12:22, 10 April 2019 (UTC)

Hi WereSpielChequers -- here’s the methodology we used for the forecasting. We took the last years data which is split by definitions of global north and south. We don’t love these terms but this is how the data was segmented.
We then applied what we saw as reasonable growth rates for various interventions split by marketing, distribution (essentially google referrals) and new product features. We used data from page previews, our SEO work and some other projects for benchmarks. We don’t have a lot of data on marketing particularly outside the US/Europe so we tried to use conservative estimates.
The 20% is essentially a blend of rapid growth in the global south and much slower growth in the global north. We want a blended number because we can’t forsake our users in the global north - after all they provide the vast majority of our community, funding and readers and our outlook must remain global. The global south growth rates are broken out and do represent significantly larger growth than in the global north but are brought back to earth by the gravity of the size of our global audience. TNegrin (WMF) (talk) 08:17, 12 April 2019‎ (UTC)
Thanks TNegrin (WMF) Good to hear this is based on forecast growth, though I suspect a cautious forecast. But the forecasting is unnecessary if we say the strategy is to increase readership faster than the internet increases its number of human users. More to the point, if at the end of five years the internet has only increased by 16% but we have increased our number of users by 18% we would probably want to count that as a success, but if the Internet grows by 30% and we only increase readership by 21% we probably want to count that target as not achieved. Forecasting percentage growth is no doubt useful to the people planning hardware purchases, so one team need to do that as well. But since we can't influence increased take up of the internet we shouldn't predicate our strategy on a particular forecast. WereSpielChequers (talk) 10:15, 16 April 2019 (UTC)
With the outright banning of Wikipedia in emerging communities, it is likely the 2030 Plan may run into issues. Geraldshields11 (talk) 14:39, 18 April 2019 (UTC)
New bans in Turkey, old ones in North Korea. Yes it is a threat, but also a stimulus to editing especially among diaspora communities. We could amend this to "growing faster than the internet, in those countries where we aren't banned". That way a ban by a country does not stop us achieving our strategy. WereSpielChequers (talk) 20:12, 18 April 2019 (UTC)

Conduct appropriate assessments of our impact on human rights[edit]

Could you elaborate on that? All other points and metrics are understandable, in my humble opinion, in the context of the "primary directive" of free knowledge, and one could be for or against if more weight should be given to one or another point. But throwing human rights is a "big word" :-D that I feel needs more context. Could you provide an example (doesn't have to be realistic or a clear commitment, this is high level plan) of what some actionable that would entail, and its relationship with free knowledge advocacy? I am not against or for it, I just not understanding it. Maybe it is an euphemism for something else concrete, but it was left vague to not commit to a specific course of action? Maybe a specific human right or set of rights is in mind? A problem that has been seen as important to fix? Apologies in advance for my lack of comprehension. --JCrespo (WMF) (talk) 17:31, 12 April 2019 (UTC)

Metrics: measuring impact of our goals rather than proxies?[edit]

I generally like the ideas in the plan, but am very concerned by the proposed metrics. In general they're proxy measurements and I'm not convinced they really represent our goals directly -- for instance while our organization's goal is to provide a system to the public to create and disseminate free knowledge in their own languages, the related metrics goals are mostly about increasing the number of "content interactions" and unique device impressions. There are a few references to language support and translation of documentation, but the rest seems to be generally subsumed into the increased interactions metric.

We've been stuck in this rut for years, since we started seeing community editorship and readership issues purely in terms of edit and user counts and of monthly interaction metrics in the hopes of making the numbers go up.

I think we need to rethink how we measure Wikimedia's success, concentrating on the actual ability of our readers to get useful knowledge and become more educated from the material in our projects. If we truly think of ourselves as a public good, then click churn is not a good metric. Not only is it trivial to game the system -- content interactions can be increased with clickbait headlines or dividing large pages into small ones -- it's just not what we're here for.

Thanks for your time! --Brion Vibber (WMF) (talk) 18:53, 12 April 2019 (UTC)

An updated Q&A with the leadership was recorded and transcribed on Friday. You can find this content on Office Wiki and there has been additional discussion around metrics here.--JUzzell (WMF) (talk) 20:13, 15 April 2019 (UTC)
+1 Vibber. i would suggest more thought from Learning and Evaluation. needs some proxies that may be harder to count, but closer to goals, which may require surveys. Slowking4 (talk) 18:32, 16 April 2019 (UTC)

Increase productivity of existing editors[edit]

I think that we should focus more on tools and technologies to increase the productivity of existing editors and to expand content faster and improve quality using automation. This can be done by substituting and augmenting human editors with tools and technologies, which allows humans to focus on more valuable edits. It's not about the number of editors and the ability to lure in and retain them, we need to improve the productivity of the existing number of editors.

  1. Use of Wikidata in Wiktionaries https://www.wikidata.org/wiki/Wikidata:Wiktionary is moving forward at speed 2. Need more resources there, it has taken years and years and not much has been implemented.
  2. Machine translation can be used originating from English language https://wikimediafoundation.org/2019/01/22/google-and-wikimedia-foundation-partner-to-increase-knowledge-equity-online/ Maybe this should be mentioned and elaborated on, because artificial intelligence is set to become a game changer before 2030.
  3. Local support and language options could be added to this helpful tool that is of great help https://tools.wmflabs.org/refill/ Maybe take it to the next level from here, because Wikipedia is collapsing under the exponentially growing problem of "error not found 404" source links and a tool is needed on every Wikipedia to fix references and external links.
  4. Guidelines about launching the extension FlaggedRevs https://www.mediawiki.org/wiki/Extension:FlaggedRevs in Wikipedias and Wiktionaries should be written about. What is the number of articles, the number of edits in local wiki per day and the size of community required to deem the launch of it a good idea and what is a good basic implementation of it in a small wiki of 400-600K articles?

I also think that we can get by without the metrics, because the reality is such that content is king and Wikipedia has no competition in local languages. So, I'd say that we need only need to know the direction we move into in terms of content, not exact numbers that have to be met in terms of page views. The uptake in developing countries often cannot be hastened by any particular website as it is more about smart phones and access to Internet. --Hartz (talk) 18:55, 12 April 2019 (UTC)

you will need to measure your productivity gains, in order to continuously improve editor productivity. quality circles, learning patterns, and training in best practices will be required.Slowking4 (talk) 22:31, 12 April 2019 (UTC)
@Hartz: Indeed, automation to expand content and improve quality is a major goal of our machine learning efforts going forward, and machine translation is one of the first footholds for us. Thank you for your suggestions, you have pinpointed many of the areas that are important to focus on, and great questions about the guidelines for FlaggedRevs, I suspect those same questions will be relevant for any extension or tool that relies on automation going forward. EBjune (WMF) (talk) 18:16, 15 April 2019 (UTC)

Markets[edit]

What do you mean by "markets"? We don't sell anything, do we? --Mautpreller (talk) 19:09, 12 April 2019 (UTC)

"Market" in English is used to refer to a geographical area or demographic group of potential customers for a service, both for for-pay and non-profit services. In this case, users or potential users of the web site "Wikipedia" and related services living in particular areas. --Brion Vibber (WMF) (talk) 19:16, 12 April 2019 (UTC)
Read it. Seven of eight meanings refer to commercial use, the eigth at least in part. Why do you use this term? The idea is not to serve the demands of customers (this is market language), the idea is to create and share knowledge ("knowing-that" as well as "knowing-how"). The question is not whether this knowledge is such that "customers" want it, the question is whether it is really knowledge. There are many commodities that "customers" like but not all of them are helpful. --Mautpreller (talk) 19:42, 12 April 2019 (UTC)
Agree completely. --Brunei (talk) 20:10, 12 April 2019 (UTC)
I'm not sure what you're trying to say. Are you simply objecting to common English terminology, or do you suggest a more substantive argument? --Brion Vibber (WMF) (talk) 20:12, 12 April 2019 (UTC)
I object to the way of thinking that is expressed in this market terminology. It's the Google way of thinking, serving "customers" in "competition". In my view, Wikimedia projects should be emancipatory and educational projects. They should offer opportunities to create and share reliable knowledge. Such knowledge is not always easy to consume and may even be disturbing and irritating. It is not advisable to tell a maximum number of "customers" what they want to hear and how they want to hear it.--Mautpreller (talk) 20:53, 12 April 2019 (UTC)
if you want to be the un-google, you will need a consensus on nomenclature. you could lift a lot of it from library sciences, but they are going to non-profit management buzzwords as well. Slowking4 (talk) 22:34, 12 April 2019 (UTC)
We mean the geographic definition, not the economic one. We are struggling a bit here with the shorthand terms, but we are generally talking about the places in the world where lots of people are coming online for the first time and where Wikipedia needs to grow to serve them better. --Lgruwell-WMF (talk) 23:50, 12 April 2019 (UTC)
@User:Lgruwell-WMF: diese Probleme könnten vermindert werden, wenn WMF das angeblich "global movment" auch als solches verstehen würde. Zu global gehören, sagen die Gerüchte, mehr als nur Englisch-Sprecher. Übersetzungen in zumindest einige große Sprachen (Sicherlich französisch, spanisch, arabisch, vielleicht Chinesisch(? fraglicher Sinn wegen Sperre!?), Russisch, Deutsch) und zwar von wirklichen Dolmetschern ... Ich weiß, es ist kaum zu glauben, aber es gibt wirklich Menschen mit klugen und wichtigem Input die kein (ausreichendes) Englisch sprechen. ... Mir ist natürlich auch klar, dass diese Vorstellung ganze Weltbilder ins wanken bringt. Aber ich wollte es trotzdem mal erwähnt haben. angeblich sind ja die Autoren so wichtig usw. (wichtig, solange sie nichts kosten, schon klar) ... Sicherlich Post 06:44, 13 April 2019 (UTC)
I completely agee. Yann (talk) 14:01, 13 April 2019 (UTC)

It's not only a question of the wording. For instance: What does it mean to "serve people better"? If you want to know what en:relativity theory really is you will have to cope with very abstract issues, no matter in which language or by which media. There is simply no way to make this a "compelling consumption experience." Any reader or recipient will have to make a considerable effort.Mautpreller (talk) 10:06, 13 April 2019 (UTC)

  • Please dear colleagues, don't you understand that "Market" is a concept that goes beyond the economic spectrum, and which can be used to define any potential interaction between supply and demand. And even if we talk about free educational content. That is a somewhat unfair trial IMO. Christian Ferrer (talk) 14:22, 13 April 2019 (UTC)
I tried several times to explain that the problem is not only the wording but the concept. Don't you understand? I try to put it as simply as possible: Recipients may look up Wikipedia because they want to be confirmed in their respective world views, with the least effort necessary. You can try to conform to this "knowledge consumption mode". This could be called a market-oriented approach: there are many people who want this, so let's do it. But I say this is wrong because this is not knowledge. Knowledge is invariably something that needs effort. Knowledge is not always comfortable and easy to create, acquire, and process. More often than not, it will be confusing, irritating, and cumbersome. Knowledge acqusition (and creation!) may well force you to question your world views. Below, you see that user:Brunei observes that reliability of knowledge doesn't play any part in this plan. I think this is symptomatic. I think that one reason for this is the concept of market. --Mautpreller (talk) 16:36, 13 April 2019 (UTC)
you have a constructivist philosophy of knowledge acquisition. but that has very little to do with the sum of all knowledge. we will need to collaborate with many editors with various philosophies; dictating one right way will be counter productive.Slowking4 (talk) 19:08, 15 April 2019 (UTC)


The team has been considering the use of the word “market” in the strategy, and while I understand your concern in terms of it being business focused, we decided to stay aligned with a term that represents the verbage used by the global world. We found the business corporations, non-profits, and world leading organizations like the United Nations all refer to this emerging “market” footprint. We remain open to suggestions on how we are able to ensure a cross-cultural and cross-sector understanding with this terminology. Let me know if you have any thoughts.--JUzzell (WMF) (talk) 20:26, 15 April 2019 (UTC)
Like all the others, I agree with Mautpreller. This terminology is redundant and unhelpful. I'm sure you can easily find replacements if only you abandon the false premise that you must do what "everyone else" is doing or that what you see close to you is "the global world". Nemo 20:06, 16 April 2019 (UTC)
i do not agree. i find the quibbling about terminology to be counter productive. dictating the word choice of the WMF is not collaboration. Slowking4 (talk) 15:33, 18 April 2019 (UTC)

Platform evolution[edit]

I am not interested at all in "machines creating content" or places where "AI is creating content". I have been a steady contributor to German Wikipedia for fourteen years. As soon as "content" is "created" by machines or AI in my home Wikipedia, I will leave at once. This would be exactly the opposite of an emancipatory project, let alone an education and knowledge project.--Mautpreller (talk) 19:25, 12 April 2019 (UTC)

Well said. Bianchi-Bihan (talk) 19:56, 12 April 2019 (UTC)
+1 from 7-year Ukrainian PhD in biophysics contributor. --Brunei (talk) 20:09, 12 April 2019 (UTC)
I believe most AI/ML proposals are tools to help contributors classify and prioritize material to work on. I think that would be useful, given that poor support tooling for repetitive editor tasks has been listed in other feedback as a common problem. --Brion Vibber (WMF) (talk) 20:11, 12 April 2019 (UTC)
Why, then, does the proposal talk about "creating content"? I am not against assistence systems that help humans to "create content" but this is not what the text says.--Mautpreller (talk) 20:42, 12 April 2019 (UTC)
My understanding is that "content creation with AI" is something to expect in less mature wikis with substantial content gaps. It could indeed be helpful to increase readership in regions that are currently underrepresented (Asia + Africa, as mentioned). However, if there is already AI content available it might be difficult to build a (human) community on top of it. —MisterSynergy (talk) 21:42, 12 April 2019 (UTC)
Interesting. So you are also somewhat skeptical concerning this feature. I think that there is a de:Zielkonflikt, i.e. conflicting goals. Is it more important to increase readership or is it more important to build human communities? My priority is building human communities for creating, organizing and sharing human knowledge. Mautpreller (talk) 09:57, 13 April 2019 (UTC)
Well, I try not to judge that much about "strategic plans" such as this one; I try to describe what I take from the proposal and which potential implications I see, positive as well as negative ones. Here, there is a "Henne-Ei-Problem" (chicken-and-egg problem) for less mature wikis which roughly reads: no content --> no readers --> no editors --> no content. One could of course just wait for underrepresented "markets" (yes, I have seen the other topic further up) to gain momentum by themselves just by organic growth, but since Wikipedia in its entirety is not very fresh any longer, I am not optimistic that this would yield satisfactory results even until 2030. The only factor one can actively boost in the Henne-Ei-problem is the content, as (basic) knowledge is already available somewhere within Wikimedia, but maybe just not available in the natural language of a particular underrepresented "market" or wiki project. Automatic content creation is just about that. However, when doing so it is crucial to have some incentives which transform the potential new readers into new editors as well. I am sure that they (WMF) are aware of this problem, but I do not know whether there is already an idea how to go about it. —MisterSynergy (talk) 18:13, 13 April 2019 (UTC)
I look at it from another point of view. The unique quality of Wikipedia is that you can do something, that you are able to work upon a representation of knowledge yourself. There is no formal barreer, no one demands your credentials as an approved expert, as soon as you edit the content is there for everyone to see. As we all know, this is not only a great opportunity but also a problem. But the fascinating and attractive "user experience" that was (and, in part, still is) made in Wiki projects was that you can do your own thing in areas that were previously reserved to certain formally qualified groups. This not only applies to writing articles ("content") but also to the creation of a common framework: discussion, self-organization, institutions. In this perspective, I think that borrowing or adopting contents from other Wikipedias is not a way to create such a user experience since the most important thing is missing: the quality of free self-making. You could think of colonialization: people learned useful things via colonialization, no doubt. But they did not learn to do their own thing. This is only possible under conditions of autonomy. Automated content creation is an even more heteronomous form of learning. You can criticize persons, you can rebel against them, you can quarrel and fight. You cannot quarrel with an automated process. Social interaction with it is impossible. To be sure, you can learn something by it. This is never bad. But it is not a way to autonomy.--Mautpreller (talk) 20:00, 13 April 2019 (UTC)
@Mautpreller: Brion is correct, as far as our plans going into the future are concerned, we are going to be looking for ways to help editors close content gaps using machine learning tools. However, we are also of the firm belief that there should always be a human at the end of any pipeline that leads to suggestions. EBjune (WMF) (talk) 18:00, 15 April 2019 (UTC)

Mit meinem bisschen an Erfahrung von Wikidata hab ich das Gefühl, dass gar nicht so wenige glücklich damit sind, Berge an Daten gespeichert zu haben, solange die Berge riesig sind und weiter wachsen, ungeachtet dessen, ob (oder eigentlich: dass) da viel Müll dabei ist, den man weder per Bot noch per Hand je aufarbeiten kann. Das Tolle an Bots ist ja, dass sie in Sekunden mehr Content "schaffen" können, als Dutzende Freiwillige in Monaten sorgfältig nachgearbeitet kriegen. Wikidata hätte wahnsinniges Potential, wenn denn auch Datenqualität und Redundanzvermeidung auf irgendeiner Agenda in irgendeiner Chefetage stünden. Parallel dazu werden Wikipedien mit Botartikeln geflutet, wo es zuguterletzt pro im Projekt aktivem Menschen hochgerechnet eine Million Artikel gibt, die in guten Teilen aus halbseidenen Quellen-Datenbanken zusammengeneriert worden sind. AI ist ja per se mal nichts Schlechtes, aber oft ist da deutlich mehr A drin als I. Man verzeihe, dass ich es vorziehe, dies in einer Sprache zu schreiben, in der ich mich auf muttersprachlichem Niveau ausdrücken kann. → «« Man77 »» [de] 13:28, 13 April 2019 (UTC)

  • +1 (as a contributor of mainly WD) --Marsupium (talk) 10:40, 14 April 2019 (UTC)
Via Google Translate (note that this may have to be redone properly by a native speaker): Jc86035 (talk) 11:23, 16 April 2019 (UTC)

With my little bit of experience of Wikidata I have the feeling that not so few [i.e. many] are happy to have mountains of data stored as long as the mountains are huge and continue to grow, regardless of whether (or actually: that) there is a lot of garbage [with it], which you can [never catch up with] neither by bot nor by hand. The great thing about bots is that they can "create" more content in seconds, [when] dozens of volunteers [over] months are carefully [reworking] [that content]. Wikidata would have tremendous potential if data quality and redundancy avoidance were on any agenda on any executive level. At the same time, Wikipedias are being flooded with bot articles, where, at the very least, there are a million articles extrapolated per person active in [each] project, [and the articles] have been compiled in good part from [dubiously sourced databases]. AI is not a bad thing per se, but there is often a lot more A in it than I. [Forgive me] that I prefer to write this in a language in which I can express myself on a mother-tongue level.

If I understand this correctly, I do agree with this, although Wikidata is still in a relatively early stage of its development cycle and it's understandable that making the data perfect is not yet a major concern (although I have noticed some fairly egregious and obvious errors in certain areas). However, kicking the can down the road is likely to result in more errors which have to be corrected at a later date. The DBpedia grant will probably be beneficial in this regard, as well as other (non-neural-network-based) initiatives like WikiCite and constraint violations development. Jc86035 (talk) 11:23, 16 April 2019 (UTC)

"integrating content from Commons, Wikidata, Wikisource and other projects into Wikipedia. This will be accompanied by rich authoring tools and content creation mechanisms for editors that build upon new capabilities in AI-based content generation, structured data, and rich media to augment the article format with new, dynamic knowledge experiences." So data will become king? As long as it is structured that is all that is going to matter? Whether the data is accurate or not won't matter as long as it is structured. This is a no-goer until Wikidata content, for one, is verified and not just the complete mish-mash that it currently is - of often unverified content made ever more complicated to understand by a user interface that allows for no commenting or explanation. Extracting structured data from Wikipedia content is one thing but making Wikipedia content based on structured data is poor thinking. Nthep (talk) 15:00, 13 April 2019 (UTC)

Wikimedia is not just Wikipedia[edit]

Hi, We should give more importance to other Wikimedia projects. I believe it is now easier for new comers to edit Wikisource or Wikimedia Commons or Wikivoyage than Wikipedia. It is easier to find in these projects areas with little or no coverage, and they also require lesser technical knowledge: uploading a picture in Commons, proofreading a page in Wikisource, or creating or updating a page in Wikivoyage is easier than editing or starting an article in Wikipedia, specially in the biggest Wikipedia projects. So we should promote these projects first. Once these new editors will have some experience, they can also edit Wikipedia. Regards, Yann (talk) 19:48, 12 April 2019 (UTC)

I think that's actually an interesting point — it's harder to learn how to edit on WP than many other wiki sites, which should probably be considered when encouraging new editors. Wikivoyage can be challenging for newcomers since there are a lot of formatting rules and customs; I would think the easiest to learn is Wikiversity. However, Wikivoyage has an arrivals lounge where new editors can get help. There are also "edit listing" buttons, etc., that make just adding a paragraph or a restaurant's phone number somewhat easier. SelfieCity (talk) 14:23, 14 April 2019 (UTC)
We have discussed the full portfolio of projects as a part of this medium term plan. The plan is iterative and will certainly impact the other projects over time. As we steadily improve the gateway to Wikipedia, we are certain that the other projects that are also important knowledge platforms will also grow in visibility and ultimately newcomer growth.--JUzzell (WMF) (talk) 20:16, 15 April 2019 (UTC)
@Yann: Wikisource is not simple to edit because of having many subpages (the same challenge for newcomers in Wikibooks) but I agree with you in promoting other Wikimedia projects and we need to find a way for removing challenges for newcomers in these projects. --Doostdar (talk) 06:56, 16 April 2019 (UTC)
wikisource is harder to edit, because the visual editor roll-out was nice but retains wikipedia menus that do not apply, thus requiring wikicode knowledge. wikipedia with visual editor is much easier; new editors are interested in articles, they are not interested in other projects, except as they support articles. Slowking4 (talk) 18:19, 16 April 2019 (UTC)

Where is reliability?[edit]

In strategic directions of the movement there are 6 statements containing "reliable/reliability". Here I find only one mentioned. Does it mean we lost from view our important goal in the world of post-truth and propaganda? What will we do if in 1-2 years nets of political trolls will invade one of the medium Wikipedias (i. e. Hungarian or Urdu) fill it with lies, get trained and then capture bigger projects? Are we ready for this? What is our plan? --Brunei (talk) 20:24, 12 April 2019 (UTC)

I wonder about this. I also read this document as a content creator from the Global North and don't see myself represented in the medium term - that the goal is AI and the Global South. I'm not sure that these bets are wrong if I were allocating resources as the Foundation (in fairness the Global South isn't on a bet for replacing the North, just growing the movement), but it's still a touch pill to swallow to see that perhaps I won't have a place in the movement. Barkeep49 (talk) 05:20, 13 April 2019 (UTC)
I agree on these points, although I actually didn't notice that they had neglected to mention reliability more than once. As I noted below, there are probably socioeconomic and possibly ethical issues with deliberately trying to drive contributor growth in the Global South, and it also seems to ignore the reality that those in the Global North are still on average in a much better position to contribute in the first place. Jc86035 (talk) 17:52, 14 April 2019 (UTC)
I think on these wikis reliability is very important. I know people who criticize Wikipedia for not being accurate, and that's something that's really going to lose editors, perhaps more than almost anything else. However, I'm not sure what measures could be put in place to make content more reliable and accurate — there's recent changes patrol, but no system for fact-checking recent changes. SelfieCity (talk) 14:25, 14 April 2019 (UTC)
While some of the wording within the medium term plan may be slightly different than that of the strategy, the medium term represents the priorities and deliverables that fully support the strategy and they work collectively. --JUzzell (WMF) (talk) 20:17, 15 April 2019 (UTC)

"We will have strong and empowered movement leaders and affiliates"[edit]

Leider hat diese Formulierung für mich, insbesonders als Bürger der Bundesrepublik Deutschland, einen negativen Beigeschmack. Apart from this, the content that is written under this section isn't written in the spirit of flat organization in many wikis. But it really good point out the move that is hapened now: Many users, that do really good contrbutions, are not able to undestand what happens with their works. This isn't only a move in the wiki or internet world, it's everywhere. The question is how we could for an understanding of this world. We do this, we stand for fredom, real-facts and transperency. And that's what we should support. If nobody knows how the wikimedia works, there are two options: we animate them to contribute with nudging or with understanding. And here we should support for that what we stand for, freedom of knowledge. Habitator terrae (talk) 21:41, 12 April 2019 (UTC)

@Habitator: Thank you for this input on how the idea of "movement leaders" comes across in the document. The idea here is not to refer to, or suggest hierarchical, top-down leaders but rather, to build the capacity of those who play a range of roles from admins, bureaucrats, stewards, affiliates, functionaries, technical contributors, and other movement organizers. The goal is to make sure that there are clear paths for people to contribute and facilitate the contributions of others, and to do so within the spirit of a flat organization, as you say.--ValerieDcosta (talk) 18:32, 16 April 2019 (UTC)
i strongly endorse the training and development of movement leaders. you need to invest your abundant monetary resources, in your scarce human resources. but the 5% editor growth will be difficult; the larger wikis are in an editor plateau, which will not grow at that rate, without significant culture change, and healthier community. Slowking4 (talk) 22:41, 12 April 2019 (UTC)
The question is, what is a "movement leader"? Searching in Meta-Wiki does not help me, finding an answer: [1]. There I see only some Wikimedia-Organizations declaring themselves as "movement leaders". When using labels like this, there should be some background information, what they mean. So that the Medium-term plans can also understand by people from other countries (if the WMF is not interested to undertake the duty of translation). In German, the direct translation from "leader" is en:Führer, and I hope You can understand, why this does not raise good associations for German speakers. I am not working in Wikipedia, because I want to be part of an ideological "movement" or because I want follow self-declared "leaders". --Magiers (talk) 12:05, 13 April 2019 (UTC)
I'd also like to know what you mean by "leaders." Who is a leader? Jimmy Wales? Katherine Maher? The Board? Admins? I always took for granted that communities should not be "led" by any of these persons, that they should be autonomous. Maybe sometimes "leadership skills" in certain areas could be desirable, but this does not define a role of a "leader". Maybe "weak leaders" might be a better goal: people who are able to inspire others but who are not in a position to enforce their ideas upon the communities. --Mautpreller (talk) 12:23, 13 April 2019 (UTC)
I think what is called as leaders here should be more something like mentors or speakers. But the word leader dose not affiliate with a democratic structure for me. --GPSLeo (talk) 12:47, 13 April 2019 (UTC)
we need skilled leaders who are capable of using the "facilitating leadership styles". we have an abundance of weak leaders, and toxic leaders. for the myth of leader-less movements, see also Jo Freeman The Tyranny of Structurelessness -- Slowking4 (talk) 22:18, 13 April 2019 (UTC)
This means you dislike the current culture in the "movement" and want a culture change towards more and stronger leadership. Does the WMF want this also? And if so, when and based on which evaluations this decision was taken? In the strategy direction "leading" was only used for the "we" of the complete movement. So I am a bit surprised, this has changed now into putting selected people "into positions of strategic influence in the movement" to lead the rest. --Magiers (talk) 23:49, 13 April 2019 (UTC)
I agree that leadership skills could be useful but definitely not leaders in the sense of command & control. We have seen far too much of this.--Mautpreller (talk) 09:54, 14 April 2019 (UTC)
Of course skills could be useful, also but not only leadership skills. It would be good to empower the movement on many levels in many ways. I just wonder, why this focus on the "movement leaders", on "positions of strategic influence in the movement". This shows to me a hierarchic thinking, that may be suitable to the WMF as an organization, but not to the so called "movement", that is build on autonomy, flat hierachies and free participation of volunteers, that don't work in the projects, because they look for someone "to rule them all". --Magiers (talk) 10:06, 14 April 2019 (UTC)
i do dislike the current culture in the "movement", but want a culture change away from toxic leadership towards competent leadership. we have a current command and control culture: "do what i tell you, or i will block you"; "change the name or i will veto the proposal". training in the "facilitation leadership style" will be required. [2], [3], [4] "This shows to me a hierarchic thinking"; no, it reflects the nomenclature of the academic literature, not a mindset: not acting because you do not like the nomenclature in the literature, is dysfunctional. Slowking4 (talk) 18:16, 15 April 2019 (UTC)
No a movement dose not need leaders. A movement needs people with special skills as speakers and people for mediation and moderation but a movement dose not need a leader. Only groups where the group members do not work on the same vision of a better world need a leader but then this is not a movement anymore. Our vision is to make knowledge accessible and improvabble for every human on the way to this utopia we will have many discussions they need moderation and in some cases a mediation between different groups. (And of course we need technical administration too) But that is nothing I would describe with the word "leader" or "leadership skills". --GPSLeo (talk) 19:17, 17 April 2019 (UTC)
you have an ideology divorced from reality, and literature. "The centrality of communication and the need for continual organization in a social movement makes their leaders particularly imporant."[5]; "Ganz begins by noting that “Social movements emerge as a result of the efforts of purposeful actors (individuals, organizations) to assert new public values, form new relationships in those values, and mobilize the political, economic, and cultural power to translate those values into action”" [6] we might agree that there are no leaders on Commons, but for me that is a bug, not a feature.
see also "Rules for Radicals" by Saul Alinsky. as Jo Freeman discussed 40 years ago, "non-leader movements" result in "secret leadership" by cliques. you will not be able to get a consensus for "a movement dose not need leaders" here. Slowking4 (talk) 15:57, 18 April 2019 (UTC)
An ideology must not be based on literature! But on the reality (which is for every person different). And my ideologie is a democratic: That means not, that a leader couldn't be democratic legitimed. But a democratic structure only works, we see in the "bug"-history of Wikipedia, if all are able to be involved. And there a secret or an unsecret leader does not help, because every is a secret for others, that aren't involved. So we must see on the realities of many person don't be involved: For them it would cost to mush time to understand the whole. Many would say the "we need a strong and empowered leader", but that isn't missing, because Wikimedia stand for freedom not for leading: We need more skills to make all "strong and empowered"! Habitator terrae (talk) 17:53, 18 April 2019 (UTC)
@Habitator terrae:We mean all contributors who lead complicated long-term projects, or who run editathons, or who have positions of trust (advanced user-rights) within the communities, or who mediate disputes well, or who simply provide a good role-model in other areas. All of these examples could be thought of as "leaders". It was a word used to encompass all of those roles. I do understand that it's not perfect, but naming things is hard.--VDcosta (WMF) (talk) 21:23, 18 April 2019 (UTC)
@VDcosta (WMF): Yes, that's always a big problem, exspecially if this is an interlangual project (we could see that all users written critical about this speak German). So I hope, in the revise it won't be so much about strategic influence or other types of "power" for a group, but more about for example organisators or mentors that give other peoples the opportunity to understand, participate and contribute. Habitator terrae (talk) 21:56, 18 April 2019 (UTC)

Medium ist von wann bis wann?[edit]

Konnte keine Zeitraum finden. Medium wovon? Ich hoffe nicht von 2019 zutrauen würde ich es WMF aber 😂 ...Sicherlich Post 06:35, 13 April 2019 (UTC)

Es geht um die nächsten rund drei bis fünf Jahre, siehe auch Wikimedia Foundation Medium-term plan 2019/Process. Das deckt sich (ungefähr) auch mit den umseitig genannten Zeitangaben, wann das Erreichen der anvisierten Ziele gemessen werden soll. —MisterSynergy (talk) 09:55, 13 April 2019 (UTC)
Danke! ... Hehe, und dafür gibts zwei Wochen Zeit für Feedback? 😂 - nein, hier soll Sicherlich niemand veräppelt werden! Wir nehmen Euren Input sehr ernst ... in your dreams 🙂🙃🙂 ...Sicherlich Post 11:18, 13 April 2019 (UTC)

┌─────────┘
Via Google Translate (section header is "Medium is from when to when?"): Jc86035 (talk) 11:00, 16 April 2019 (UTC)

Could not find a [time] period. Medium of what? I hope [it's] not [the medium of] 2019 I would trust [the] WMF but 😂

It's about the next three to five years; see also Wikimedia Foundation Medium-term plan 2019/Process. This coincides (approximately) with the times mentioned overleaf, when the achievement of the intended goals is to be measured.
Thank you! ... Hehe, and there are two weeks for feedback? 😂 – no, nobody should be [kidding me] here! We take your input very seriously ... in your dreams 🙂🙃🙂

Jargon[edit]

Unfortunately the high level of techno-cryptic jargon makes this unreadable by unpaid volunteers, or in truth any real person, and near meaningless as a "plan":

  • "catalyzing convenorship"
  • "gender-equitable pipelines"
  • "paradigm of free knowledge"
  • "positive and unified free knowledge policy agenda"

These jargon packed phrases raise more questions than they are worth. It is telling that the most basic words underpinning this document have no specific or measurable definition, i.e. "markets", "brand", "movement", "platform". If the expensive consultants putting this together do not appreciate the value of plain English or why a plan has to be measurable rather than rhetoric and concepts, then they are the wrong consultants for this job.

My feedback, this is not a plan. -- (talk) 12:22, 13 April 2019 (UTC)

I fully agree. Who did put this plan together, by the way? You say that it was drafted by "expensive consultants". Who are they? Reading the plan, I thought (maybe erroneously) that this was done by some WMF groups or institutions. The first word of the text is "we" but there is no hint who could be this "we". -- Mautpreller (talk) 12:33, 13 April 2019 (UTC)
It seems to be a secret, but possibly this was 'facilitated' by the team of consultants from Wolff Olins who were the recent "brand consultancy". I doubt they were anything less than "very expensive", but due to the WMF being fully committed to transparency in a way that has yet to be defined, we will probably never know how much they cost. -- (talk) 12:47, 13 April 2019 (UTC)
This plan has been written by Foundation staff. We haven't paid any consultants to produce it. Quiddity (WMF) (talk) 04:53, 14 April 2019 (UTC)
Could you please elaborate? "We are pleased to announce" - we who? I am not interested in names of persons but in names of groups or committees.
Moreover: I understand that this draft is published for the sole purpose of community comment. Why, then, is it written in an esoteric jargon? --Mautpreller (talk) 09:47, 14 April 2019 (UTC)
I presume that there were consultants to facilitate, or were WMF employees really just put in a room and told to write a top level plan, without this being part of the consultancy driven 3 year long WMF strategy development process? -- (talk) 08:08, 15 April 2019 (UTC)
@: According to this, they started with a top-down process, then focus-grouped a demographically and organizationally representative distribution of employees to figure out some of the details, and then gave the details back to the "executive team" for homogenization. At least based on that page, only employees were involved. Jc86035 (talk) 13:46, 15 April 2019 (UTC)
Bizarre. The WMF invests literally millions in a Strategic Direction process, then has its employees go on a tangent to create a "medium term" which has no specific relationship to all that investment and apparently by design contains nothing measurable so nobody could ever use this "plan" to hold anything to account. Almost the definition of spinning wheels for the sake of it. It's reminiscent of a political manifesto that is produced just before an election, then thrown away and deliberately forgotten immediately afterwards. It would be interesting to go back to the 2018 plan, and find any independent report that holds the WMF properly to account against it, beyond subjective "I think we did all right" reports without any hard, verifiable, measurements. -- (talk) 13:54, 15 April 2019 (UTC)
@Quiddity (WMF): For clarity, are community members expected to make any edits to the draft? None have done so yet, although I thought it would be worthwhile to ask. Jc86035 (talk) 13:46, 15 April 2019 (UTC)
@Jc86035: No, because changes need to be approved by the executive team. Thanks for asking, and the minor "to" change you made to the preamble. Quiddity (WMF) (talk) 17:06, 15 April 2019 (UTC)
Fæ has a point. I didn't have the time to discuss the content, but such jargon is not useful. See also the "Markets" section above. Regards, Yann (talk) 13:53, 13 April 2019 (UTC)
I also agree with Fæ. This plan is full of jargon and I don't have much confidence given the recently "brand" kerfuffle. Abzeronow (talk) 16:59, 13 April 2019 (UTC)
I don't full agree with Fæ: The fact that the probable problems with the plan (or like you want to name it) are hidden by a very special jargon is right: But there are also really plans like spend time in thinking about brand or support structured data and AIs. That are, in my view, very critical points hidden in a mess of "gender-equitable pipelines", "human rights impact assessment" or "Worldwide Readership", that sounds really good. If you look deeper, you see that in this mess "design changes" are shown as "meaningful changes". Habitator terrae (talk) 22:18, 13 April 2019 (UTC)
i see the outcome goals of a plan. i would prefer to see the action steps associated with each goal, and then the resource loading for each action step. i find the denouncing the corporate speak to be counter productive. Slowking4 (talk) 22:22, 13 April 2019 (UTC)
Goals are not a plan. Goals are not a type of plan. A plan is the actionable steps you are going to take to accomplish a goal. GMGtalk 00:59, 15 April 2019 (UTC)
  • This might contain a modest paragraph's worth of an actual actionable plan. I'm not really sure why it's so difficult to write in plain English 1) what you're going to do, 2) how you're going to do it, 3) when you're going to do it by, and 4) how you're going to measure success. But whoever the target audience is for this, it isn't us, because the majority of it is either meaningless jargon, or impenetrable jargon, but I suspect the former. GMGtalk 19:23, 14 April 2019 (UTC)
    @GreenMeansGo: I suspect that – if this was compiled by an "executive team"(?) of Foundation employees – that they might just not have realized how much jargon was in it, because they've all been working in some sort of bubble comprising mainly other employees of various non-profit organizations. I wouldn't really fault them specifically for that, since it's probably not actually their job to write strategic plans, this is just a draft, and it doesn't totally prevent us from discerning any meaning from the words. On the other hand, it is somewhat disappointing that they focus-grouped employees instead of involving community members earlier in the process (notwithstanding that some employees are active community members), and there probably isn't enough detail to really know if the outcomes are anywhere near achievable through the methods. Jc86035 (talk) 07:53, 15 April 2019 (UTC)
This was a collaborative, multi-disciplinary process across the organization, designed to incorporate our many different and varied voices. Each outcome was contributed to by: 90 Foundation staff members (27% of total staff), more than 50 hours of process and plan design by Working Group members, 18 hours of large-group collaborative strategy sessions, and dozens of hours of group and individual writing time by more than 50 people. We worked hard to ensure there were many participants and points of views in the design of this plan.--JUzzell (WMF) (talk) 20:19, 15 April 2019 (UTC)
  • That makes it a bit worse actually that you put all this time and effort into this, when it isn't really a plan at all; it's an aspiration about what your plan will eventually do once you make it. A "plan" is not 1) we have these objectives, 2) en:Black box, 3) our progress on our goal will increase by 30%. That's what most of this actually boils down to, and the percentage points seem entirely arbitrary, with no evidence based reason why they should be set where they are. The bit that goes where the black box is, is the part called the plan. GMGtalk 21:32, 15 April 2019 (UTC)
Note: see also the #Markets section above for a comment specifically about jargon. Quiddity (WMF) (talk) 20:38, 15 April 2019 (UTC)
  • As an example:
We will facilitate the smooth entry, retention, and participation of new participants and contributors
Great. And we will accomplish it by:
by providing them with safe & efficient onboarding
By doing what? What specific changes are to be made and in what way?
programs and convenorship opportunities
What programs? What opportunities?
platforms for learning, networking and mentorship
What platforms? What are they going to look like? How will they function?
The answer to these questions is the plan. This is the equivalent of convening a business meeting and saying "we will increase profits by creating programs, platforms, and opportunities that increase productivity and reduce cost". You haven't said anything other than "we will accomplish something that is intermediate to our goal". In social-work-speak, you've stated a goal, and you identified an objective, but you haven't made a plan at all.
We will build a house. How? By making a structure that is secure and interconnected. But that doesn't tell me what lumber to buy, where to pour my foundation, and where to put my hammer. GMGtalk 22:53, 15 April 2019 (UTC)
@GreenMeansGo: Thank you for the feedback. You are absolutely right, the plan is missing much of the operational details about HOW we will accomplish our goals and outcomes. That was intentional, as we wanted to keep this at the strategic level. Please see our rationale in the second paragraph under The Process. Also, in response to the rest of this thread on jargon, thank you for pointing out the phrases and terms that do not resonate with you. We will work to improve the language in the next iteration. We hope this does not deter you from continuing to provide feedback on the outcomes and metrics we've selected to focus on! See some guiding questions for feedback. Thank you! --DMccurdy (WMF) (talk) 00:12, 19 April 2019 (UTC)

Wikidata[edit]

Ok, j'espère donc qu'on aura enfin tous les moyens nécessaires pour traduire les infoboxes et tableaux Wikidata dans de très nombreuses langues. Parce que ça ne sert à rien d'avoir les données si elles ne s'affichent au mieux quand dans une version linguistiques, c'est aussi intéressant par ce biais de faire comprendre qu'un contributeur peut par ce biais contribuer à plusieurs versions linguistiques, là où de toute manière on manquerait de main-d'œuvre. Jérémy-Günther-Heinz Jähnick (talk) 17:23, 13 April 2019 (UTC)

┌──────┘
Via Google Translate: Jc86035 (talk) 10:52, 16 April 2019 (UTC)

OK, so I hope we will finally have all the means necessary to translate Wikidata infoboxes and tables in many languages. Because it is useless to have the data if they do not display [optimally] when in a [given] language version, it is also interesting in this way to understand that a contributor can in this way contribute to several language versions where, in any case, we would be short of manpower.

Paternalism and hubris[edit]

"For Wikimedia to continue to advance free knowledge on our own platforms and through the broader movement and mission, we will need to maintain and grow our readership globally. This means holding steady in places of high engagement, and increasing engagement in places of low Wikimedia awareness and use. By 2030, 80% of the world’s population will be in Asia and Africa -- regions with the today’s lowest Wikimedia readership representation. We must close this gap if we are to advance our mission of free knowledge, built for and by the world." Don't you feel that this is a paternalist attitude? "We", meaning the WMF in San Francisco, want "them" in the Global South. Fine. But maybe "they" in the Global South don't need automated translations of "our" texts but want to do their own thing? Maybe they even don't want to align to "our" values, etc.? My feeling is that the whole document is an expression of extreme en:hubris. "We" (please don't count on me!) want to be "the essential infrastructure of free knowledge". Regardless of the people who are not delighted by such a vision that defines the "Wikimedia Movement" as a monopolist. You may be sure, there are a lot of people within the "movement" who don't agree.--Mautpreller (talk) 10:22, 14 April 2019 (UTC)

Focus[edit]

Some of the things here are good, even if they might be a bit on the vague-and-confusing-marketing-speak side. However, I feel like it it's still somewhat problematic (particularly the "thriving movement" section): any contributor could have written, at this point, that it would be a good idea to try to increase editor retention and diversify editor demographics by some arbitrary percentage that sounds achievable. It's a vaguely good direction, but it doesn't really seem to address many of the practical issues which affect editors and editor retention right now.

I've written out some points which I think could be helpful, although they may be limited due to my perspective, and most of them are based on the idea that high-activity contributors may be inherently more useful than low-activity contributors solely due to being more productive. (I became a Wikimedian in 2012 and mainly edit on the English Wikipedia, Wikimedia Commons and Wikidata.)

  • Impact: The English Wikipedia still accounts for 48% of all Wikimedia traffic. A large number of WMF projects are used less than some individual English Wikipedia articles. Increasing the reach of smaller projects in the short term may help in the long term, but until those projects' content can compare in quantity and quality to that of the ten most-used Wikipedias, a majority of those who speak English as a second language will likely use the English Wikipedia article instead of or in addition to the one in their first language. For simplicity, I will mainly use the English Wikipedia as a reference point, even if other Wikimedia projects have vastly different needs; I would be lying if I were to claim that all Wikimedia projects are created equal.
  • Quality: Most of Wikipedia's good and featured articles are written by extremely-long-term editors (even I have only gotten an article to C-class).[1] Writing a good article, particularly for pre-Internet topics, often requires reading through entire books which might not be in English, and (even because of this alone) may be significantly difficult for most contributors for various reasons. This will likely be the reality until it becomes possible for an arbitrary digitized book to be fed through a neural network and turned into an encyclopedic summary. As such, increasing the number of short-term editors may not even make much of a dent in Wikipedia's genuinely good content, even if many readers only come for the lead sections and infoboxes of articles.
  • Quantity: Very few users become active long-term editors; and one long-term, high-activity editor could over time do the work of twenty long-term, low-activity editors or a thousand short-term, low-activity editors. Improving the number of editors who stay for one month would be comparatively low-benefit compared to improving the number of editors who stay for five or ten years; most of the mindshare comes from reading and not editing anyway. Personally, I think the WMF should actively seek those who would have the time, dedication, energy, resources and ability to become long-term, high-activity editors, and not just make software improvements in the hope of those productive editors coming of their own volition (making the interface better usually isn't going to fundamentally change how good articles are written, even if it increases the average amount of time that short-term editors will stay active). Furthermore, not everyone wants to edit Wikipedia, and it's possible that only very specific sorts of people become high-activity, long-term editors, so casting the net as wide as possible (which is effectively the current strategy) might not catch many more people. (This is not to say that diversity attempts are pointless, and I think they have a place in the movement strategy, but there are many outside factors which almost certainly have a significant effect on which demographic groups are more likely to edit Wikipedia and enjoy doing so.)
  • Time: It takes time and practice to get good at editing, so people who would be great editors but just do not have the time to contribute – even in the developed world – might not ever edit if they have children or work long hours. This is probably a significant factor and it can't be measured directly through editor retention rates. Employing people to edit Wikipedia is probably out of the question because of the complications and conflicts of interest involved, although compensating amateurs to take (or license) particular photos for Commons could be somewhat beneficial in the medium term (i.e. the century and a bit remaining until most of today's photos and videos are out of copyright).
  • Dedication and energy: There are only 24 hours in a day, and e.g. Netflix at least claims to compete against basically everything (e.g. Fortnite, YouTube) and not just Hulu and HBO,[2] so even those who have free time may need to be actively drawn away from other things in order to have time in their lives for editing Wikipedia. There are many, many people who use Reddit (or other social networks) several hours per day – if they don't find Wikipedia editing too stressful for them, they may well choose to edit Wikipedia instead of going on Reddit (or whatever) if they're prodded in some way. Particularly if attracting more low-intensity users is desirable, attracting people from other time-consuming things (e.g. through native advertising, by making promotional posts on the respective services) could be important.
  • Resources and ability: Most people in the developing world probably do not own laptops, even though most may, by now, own smartphones. It is probably fundamentally unfair and possibly unethical for those in the developed world – i.e. most of the present Wikimedia community and WMF staff – to ask those who can't really afford a good laptop (in addition to everything else) to spend their free time editing Wikipedia on a slow, tiny interface, probably in their second or third language. (Also, regarding Commons, consider the cost of a good camera.) The privileged and educated are more likely to be able to edit Wikipedia than the disadvantaged (and are more likely to be able to do it well), so of course there's going to be a much larger proportion of contributors coming from developed countries, and it may have to stay that way for the time being. (I am not arguing that the WMF should exclude low-income editors from contributing, but that it is questionable whether asking them to basically work for free is a good idea at all.) If it is not the WMF's job to improve external factors to increase the number and distribution of editors, then it should be accepted that these factors are basically out of the WMF's control; otherwise, the WMF could be using more of its resources to directly lobby for (or even fund) e.g. healthcare and education improvements (whether in the US or elsewhere) instead of e.g. hiring people from consulting agencies to write vague statements of purpose.
  • Return: I do not receive any direct benefits for myself other than possible personal enjoyment from coding Lua modules, writing articles, structuring Wikidata items, grandstanding in the hope of aiding the movement (to put it harshly), or participating in WMF initiatives. Even an OpenStreetMap editor can reasonably expect to receive direct benefits, as they can draw their surroundings and directly use their own work in their daily lives. The WMF can take the contributors for granted, but it would be better to actively acknowledge more of the tens of thousands of individuals who contribute, if making editing more inherently beneficial for the editors is out of the question.

The losing American presidential candidates will probably raise about the same amount of money as the WMF will this year. If the WMF donations are to continue increasing, the WMF should probably ask if it – the organization, not the projects or the users – is having as much of a real-world impact as those twenty or so individuals. (Well, maybe not this exact question, and it's not really an apples-to-apples comparison, but it's probably a good framing device nevertheless.) I know some of my points are already addressed to some degree in the draft, and it would be nice to know if others have been brought up privately but not included in the draft for any particular reasons; but there seems to be much room for the WMF to be more ambitious, more effective and more responsive. Jc86035 (talk) 11:14, 14 April 2019 (UTC)

References

  1. Of the ten most recently promoted good articles at time of writing, from Reign of Marcus Aurelius to Kiev-class destroyer, the lowest edit count of any of the primary authors is 4,364 (Kyle Peake); and a majority of the primary authors joined more than eight years ago (Sturmvogel 66 wrote two of the articles, so 5/9 authors and 6/10 articles).
  2. Spangler, Todd (17 January 2019). "Netflix Claims ‘Fortnite’ Is a Bigger Competitor Than HBO". Variety. I know the Netflix claim is probably an exaggeration, and this is noted in the linked article, but there's likely some truth to the claim.

be an independent voice[edit]

Under the section Priority: Global Advocacy is written something, like that Wikimedia will be an independent voice. That sounds really good, but there is missing how this should be done: Wikimedia get often money by big companies like Facebook, Amazon or Google (and also write positive news about this). My question not yet answered in the plan is: How Wikimedia want to show that it is an independent voice by showing, that the priorities aren't influenced by this donates? Habitator terrae (talk) 12:21, 14 April 2019 (UTC)

Vorbe în vânt[edit]

Ce prezentați dumneavoastră este o propunere de intenții care nu are la bază o metodologie. Adică, cineva din structura organizației s-a chinuit să scoată ceva frumos și plin de generozitate, fără a lua în considerare factorul uman. Cu cine vreți dumneavostră să faceți treaba? Eu provin de pe ro.wiki și sunt blocat de trei luni de niște sysopi incompetenți și orgolioși, în condițiile în care ro.wiki mai are circa 30 de contribuitori cu tot cu sysopi, care sunt 20. M-am uitat pe en.wiki și am văzut azi cu stupoare că de la ora 0.00 până la ora 9.00 numărul de contribuții era sub 100, în condițiile în care en.wiki are o bază de 1.500.000.000 de vorbitori de limba engleză. Nu cumva ați ajuns ca Adolf Hitler care în ultima sa fază, conducea armatele pe care nu le mai avea? În orice domeniu de activitate din lume există o lipsă acută de profesioniști plătiți, cum veți găsi dumneavoastră profesioniști neplătiți care să implementeze proiectul pe termen lung? Lumea se degradează, în Franța circa 30% din populație este neangajabilă, este doar un exemplu. Mie mi se pare că avem de-a face cu utopia lui H.G. Wells. Asybaris01 (talk) 12:11, 15 April 2019 (UTC)

┌──────┘
Via Google Translate (section header is "Words in the wind"): Jc86035 (talk) 14:01, 15 April 2019 (UTC)

What you present is a proposal of intentions that is not based on a methodology. I mean, somebody in the organization struggled to get something beautiful and generous without taking into account the human factor. Who do you want to do? I come from [the Romanian Wikipedia] and have been blocked for three months by some incompetent and egotistical sysops, while [the Romanian Wikipedia] has about 30 contributors with sysops, who are 20. I looked at [the English Wikipedia] and today I was astonished that from 00:00 until 9:00 the number of contributions was under 100, while [the English Wikipedia] has a base of 1,500,000,000 English speakers. Have you ever arrived as Adolf Hitler who, in his last phase, was leading the armies he did not have? In any field of activity in the world, there is an acute shortage of paid professionals, how will you find your unpaid professionals implementing the long-term project? The world is degrading, in France about 30% of the population is inoperable, it is just an example. It seems to me that we are dealing with the utopia of H.G. Wells.

Bune intenții[edit]

Plec de la premisa că un astfel de plan nu poate să fie animat decât de intenții bune. Așa cum cercetările în domeniul nuclear au dus la exploatarea unor noi surse de energie, dar și la apariția armamentului nuclear, tot așa și cercetările în domeniul informatic pot conduce la dezvoltarea unor produse ce nu sunt benefice dezvoltării societății omenești. Războiul mediatic și știrile false reprezintă doar vârful aisbergului. De principiu, apreciez ideea unui proiect la care să contribuie orice voluntar. Pe de altă parte, consider că este firesc să existe un filtru dacă se dorește realizarea unei enciclopedii în adevăratul sens al cuvântului. Apreciez la modul superlativ alinierea proiectului la noile tendințe (IA, data mining, big data). Astea fiind spuse, vă propun să aveți în vedere următoarele:

  • crearea unui cont să se facă exclusiv pe bază de identitate reală, nu cu un nume care nu are nici o semnificație. Altfel spus, să existe numele persoanei și un CV din care să rezulte vârsta, locul de domiciliu (țara/eventual localitatea) și nivelul de pregătire - școli, specializare, domeniu de activitate. Și aceste informații să fie publice!
  • drepturile de modificare să fie dependente de datele din CV, dar și de contribuțiile anterioare ale persoanelor implicate într-o eventuală discuție în contradictoriu.
  • numirea în funcții de administrare să se facă exclusiv pe bază de CV, fiind strict interzisă crearea unor grupuri de presiune prin recomandări încrucișate. Să existe o comisie care să analizeze astfel de situații și să ridice drepturile de administrare acolo unde s-au creat astfel de grupuri. Apariția unor astfel de grupuri se poate foarte ușor observa analizând situațiile de conflict și urmărind cine pe cine se sprijină la nivelul administratorilor.
  • drepturile de administrare să fie grupate pe două nivele: superior & inferior. Pentru a primi drepturi de administrare din grupul superior să se impună o condiție obligatorie ca persoana respectivă să aibă domiciliul stabil în statul căruia îi corespunde limba/proiectul. (ro. - România, en. - SUA&Anglia, etc)
  • să se treacă la eliminarea din enciclopedia Wikipedia a acelor articole care reprezintă în realitate platforme de propagandă și presiune politică, fără un suport științific/real/dovedit statistic.
  • să se acorde mai mult credit surselor de arhivă decât surselor din media (din online).
  • să se realizeze o eliminare de lungă durată (3-6 luni) a celor care produc vandalisme evidente.
  • să se încerce eliminarea posibilității de a interveni în proiect a celor care nu se autentifică deși au un cont (folosesc doar IP-uri), în special a celor ce primesc drepturi de administrare. PheonixRo (talk) 13:28, 15 April 2019 (UTC)

┌──────┘
Via Google Translate (section header is "Good intentions"): Jc86035 (talk) 13:57, 15 April 2019 (UTC)

I'm assuming that such a plan can only be animated by good intentions. Just as nuclear research has led to the exploitation of new sources of energy, but also to the emergence of nuclear weapons, so research in the field of information can lead to the development of products that are not beneficial to the development of human society. Media war and fake news are just the tip of the iceberg. In principle, I appreciate the idea of ​​a project to which any volunteer contributes. On the other hand, I think it is natural for a filter to exist if an encyclopedia is to be made in the true sense of the word. I highly appreciate the alignment of the project with the new trends (IA, data mining, big data). Having said that, I suggest that you consider the following:

  • creating an account solely on the basis of real identity, not a name that has no meaning. In other words, the name of the person and a resume showing the age, place of residence (country / eventual locality) and level of education – schools, specialization, field of activity. And this information should be public!
  • the rights to change are dependent on the data in the CV, but also on the previous contributions of the persons involved in a possible controversial discussion.
  • the appointment in administration positions should be based solely on CVs, with the creation of pressure groups by cross-references. There is a committee to analyze such situations and to raise the administration rights where such groups have been created. The emergence of such groups can be easily observed by analyzing conflict situations and tracking who is supported by the administrators.
  • administrative rights to be grouped on two levels: superior & inferior. In order to receive administrative rights from the higher group, a mandatory condition for the person to have a stable domicile in the state to which the language / project corresponds is imposed. (Romania [for rowiki], USA [and] England [for enwiki], etc)
  • to move to the removal from the Wikipedia encyclopedia of those articles that actually represent propaganda and political pressure platforms without a scientific / real / proven statistical support.
  • to grant more credit to archive sources than media sources (online).
  • to achieve a long lasting (3–6 months) disposal of those who produce obvious vandalism.
  • try to eliminate the possibility to intervene in the project for those who do not log in while having an account (using only IPs), especially those who receive administrative rights.

From the English Wikipedia perspective (since I am unfamiliar with the Romanian Wikipedia's policies and culture): The first four are almost certainly non-starters and most are addressed by w:en:Wikipedia:Perennial proposals (see also w:en:Citizendium). The latter four seem to be non-sequiturs and are either addressed by existing English Wikipedia policy, or are practical or technical impossibilities. I apologize if I have misinterpreted the user's points of view due to the translation. Jc86035 (talk) 13:57, 15 April 2019 (UTC)

Yup, according to him the admins of fa.wiki should be inhabitants of Iran, admins of zh.wiki inhabitants of China and admins of ko.wiki inhabitants of North Korea. Pretty neat way of turning Wikipedia into an instrument of propaganda. Tgeorgescu (talk) 22:44, 16 April 2019 (UTC)

Strategy[edit]

Firstly, I think that it's important for all parties involved to try to stay aware of their own biases and limitations (an example, of course, would be the jargon complained about at great length on this talk page). Of course, the onus falls more on the WMF than on the contributors, since the contributors (and most WMF employees) won't have any actual role in writing this strategic plan.

As noted by GreenMeansGo above, there is not much value in writing a plan that is too high-level and ends up only describing the vague structure of a more detailed plan. I'm inclined to agree with him, even if I think that the plan does point in the right direction. I perhaps would go further, and say that a strategic plan, especially one of a large, stable organization like the WMF, should probably try to do some things differently – or at least be discernibly different – to previous plans of the same scope (whether or not they have succeeded). Otherwise, I think there's not much of a point in specifically pursuing the new strategy (why not reuse an old one, instead of spending hours on the details just to end up with something that looks pretty much like the old plan?).

I also think the WMF would carry on just fine without having any medium-term plan at all. Even if the WMF ends up having negative impacts on communities, as has occurred in the past, there's basically no correlation between community tension and donations, or between community tension and reader numbers. Aside from the development of donation banners, it would be plausible to attribute increases in both donations and readership entirely to outside factors like improvements in Internet access and increases in overall sociopolitical awareness. Perhaps the plan may not have any impact at all. Even in the event of a sudden dissolution of the WMF, the projects themselves would probably carry on regardless. Jc86035 (talk) 06:56, 16 April 2019 (UTC)

I agree with the sentiment that the plan is very high-level. If the path to achieving some of the strategic goals isn't clear, then the process of figuring that out can be a useful goal itself. – Ajraddatz (talk) 01:08, 19 April 2019 (UTC)
Sometimes the plan, either the 2019 or 2030 plan, has to be high-level; then it is up to the Working Groups to flesh it out. Geraldshields11 (talk) 22:15, 6 May 2019 (UTC)

Delivery and timing[edit]

While I, too, find this plan to abuse wooden language, after reading it several times (and connecting the dots using comments in this talk page) I think that overall the plan seems to go in the right direction. It also clarifies some things that worried me in the 2030 strategy. However, I do have some punctual concerns:

  • Metrics are not always SMART, specifically the ones in the "Brand Awareness" and "Global Awareness" section. Especially the latter seems to me that it could be better defined, especially wrt the markets that you plan to work in.
  • I am worried about pushing this plan (especially the branding part) before the strategic planning process is done. We might end up with 2 very different ideas of how to communicate our community to outsiders.
  • Do you plan to change anything in the way features are prioritised and delivered? All the projects I have followed lately (from the SVG translator to Structured commons) have seen delays. This indicates there might be a problem in the way the Engineering department is preparing these projects. Also, added up delays have the ability to derail any multi-year plan. However, the plan only talks about code reviews and customer satisfaction.
  • You seem to be using "Emerging markets" and "Underrepresented markets" more or less interchangeably. Are they the same thing? What countries go into each bucket?
  • It would be great to see examples of projects already started in each of the directions you describe. That would help clarify that this plan is not a revolution, but an evolution and would counter-balance the jargon so heavily used throughout the document. These examples can be in an annex in order to keep the plan "clean".--Strainu (talk) 13:50, 16 April 2019 (UTC)
Hello Strainu. I'll just answer on the part around the Medium Term plan (MTP) and the Movement process interacting / waiting one for the other. I can imagine that specific pieces of some of the priorities will greatly be impacted by the outcomes of the Movement strategy, and I want to confirm that the MTP is definitely "listening" to the Movement Strategy. The main reason behind a high level, medium-term plan was to allow the flexibility that a one-year only planning process doesn't. As such, while the MTP itself has some iterations points built in, at which it will evolve to take into account changes in the environment, it is important to note that the Foundation also has a number of staff members in the Movement strategy working groups, who can report in real time. We are also working on making sure that the Foundation as a whole is taking into account the visible trends that come out of movement strategy (for example looking at scope and scoping documents, listening to general community conversations etc.). We are being very careful not to preempt anything while being aware that these processes had to run in parallel, and we did not want to push back structuring the Foundation mid-term thinking and planning. As a matter of fact, the MTP is thought as a flexible enough framewrok to accommodate the structural recommendations that will come out of the Movement Strategy Process. Delphine (WMF) (talk) 10:52, 17 April 2019 (UTC)

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

Comments after 20 April[edit]

Bune intenții[edit]

In reply to #Bune intenții
We have a number of admins on EN wiki who are either English speakers living outside the English speaking world, or who have English as a second language and live in their home countries. You do need to have very good English skills to pass RFA, but we certainly don't rule out candidates not currently living in English Speaking countries. Not can I think of any logical reason to do so, except if as Tgeorgescu says, you are some dictator wanting to take control of a language version of Wikipedia. Reducing vandalism blocks from indefinite to just 3-6 months does not sound like a well thought out or logical change. As for banning IP editing, we have a theory that the easier we make it to edit the more likely people are to make their first edit, making that first edit more difficult doesn't dissuade vandals as much as it does the casually interested who are just trying to help. On that theory banning IP editing would lose us more good editors than bad. WereSpielChequers (talk) 13:33, 23 April 2019 (UTC)

Priority: Global Advocacy[edit]

Protection of honor and dignity (Global Advocacy):

  • Wikimedia Users
  • Wikimedia Organizations
  • Any people (non Wikimedia Users) Yes check.svg Done WP:LIVE
  • Other (any) Organizations
  • Respect for the laws of all countries Yes check.svg Done (Ex: Wikimedia Commons)

(in the external environment and also in all wiki-projects). — Niklitov (talk) 21:10, 7 May 2019 (UTC)

@Niklitov: "Respect for the laws of all countries Yes check.svg Done (Ex: Wikimedia Commons)": en:Censorship_of_Wikipedia#China? Habitator terrae (talk) 15:37, 10 May 2019 (UTC)

Oral Histories[edit]

By 2025, 5 global communities will participate in a pilot program to increase the presence and acceptance of oral knowledge on the Wikimedia projects.

This goal is likely to fail.

The Foundation has repeatedly advanced this idea in the past. The community has repeatedly responded with hostility and derision. Editors get rather concerned when the Foundation clearly doesn't understand what we do, how we work, and why. There is no way any functional Wikipedia will accept an "Oral History" about how Randy-from-Boisie's grandfather was abducted by aliens. I know that's not how the Foundation envisions this content, but that is essentially the content that you are proposing that we deal with. Anyone trying to add Oral Histories at a functional Wikipedia will have their content rapidly deleted, they will likely receive a swift warnings, and if they persist in trying to add this kind of content we will block them from editing.

I don't know what non-Wikipedia projects you expect to get on board, but among Wikipedias I expect your only prospect would be to try to find some microwikis where "consensus" might constitute a few inexperienced editors with little idea what they're doing yet. Please don't seek to damage and victimize the smallest most vulnerable wikis in the name of someone's pet project. Alsee (talk) 20:10, 16 May 2019 (UTC)

I'd like to disagree with this perspective and hope that the Foundation continues to work on this idea. Entire cultures, especially those indigenous to North America, rely on oral knowledge to teach and learn their collective histories. It might be a balancing act figuring out how to integrate that knowledge into Wikipedia (as indeed it is to figure out how that knowledge interacts with written knowledge in legal and general contexts), but it is still worth thinking about and doing. – Ajraddatz (talk) 20:25, 16 May 2019 (UTC)
Ajraddatz if I recall correctly, I think I've seen this come up at least three times before. And if I recall correctly, each and every time it resulted in a manifest-consensus that it was a stupid and non-viable idea. I don't doubt the good intentions behind it. However I do not believe any of the advocates of this idea have yet answered what is supposed to happen when my buddies and I start uploading a pile of fictional and/or contradictory "oral histories".
Each time this proposal comes up, the answer is the same. This is a job for reputable researchers and academic institutions and news institutions and publishing institutions. Wikipedia can and will incorporate information about&from oral histories only after it has been published in Reliable Sources. Alsee (talk) 22:02, 16 May 2019 (UTC)
@Ajraddatz: I don't think it's Wikipedia's job to record oral histories. Doing so would clearly conflict with how most Wikipedia projects make sure information is at least somewhat accurate; and in a vacuum ignoring how different cultures keep their histories, it's not very different from recording a random story of dubious origin. If the Foundation is interested in collecting this information it would probably be more valuable for a professional (probably an anthropologist) to write a book or publish a study about it. There are obviously some issues with having the publishing barrier at all, but Wikipedia is ultimately a secondary source, and it would be considered a complete waste of time to, say, collect a bunch of stories from Reddit threads and leave the verification completely up to the reader. Jc86035 (talk) 10:33, 19 May 2019 (UTC)
Equating the social and cultural histories of entire cultures to stories from Reddit is hugely irresponsible at best. I'm not saying that it will be an easy fit, I'm saying that thought needs to be put into it and there is a place for it in strategic planning. – Ajraddatz (talk) 20:29, 19 May 2019 (UTC)
@Ajraddatz: While it's probably not the best comparison, a whole lot of culture has happened on Reddit, even if it's been massively globalized and homogenized and has occurred over little more than a decade. Reddit's culture is much better catalogued and is much more heavily intertwined with other parts of (the?) "global"/Internet culture, of course; but the only realistic way I can see the WMF accomplishing this particular goal is through hiring professionals to catalogue it in other publications first. Unless they come up with something better, of course.
I don't mean to diminish the importance of indigenous cultures, of course, and I apologize if my previous comment was insensitive, but most Wikipedia editions are ultimately and necessarily dependent on the existence of today's publishing systems because of their utility as a filter for false or irrelevant information. Even if this inadvertently forms a barrier towards including information about indigenous cultures, and even if it's epistemologically somewhat dubious for Wikipedia to rely on formal publication as a substitute/proxy for correct and verifiable information, it would be somewhat unwise – and quite probably fruitless – for the WMF to rely on changes in local project policy to accommodate this information. This is really the core issue with the current wording of the plan: it probably won't get anywhere, even if this is something that could improve Wikipedia. (In general, if something's success is completely dependent on changes in established local project policy, it's probably likely to fail, regardless of whether or not it should.) Jc86035 (talk) 19:14, 20 May 2019 (UTC)
I suppose you could add the information to Wikibooks, but this would necessarily preclude its inclusion in Wikipedia. Jc86035 (talk) 19:29, 20 May 2019 (UTC)

The goal can arbitrarily be judged reached or not anyway, since "global communities" and "participate in a pilot" mean next to nothing. Nemo 13:16, 17 May 2019 (UTC)

Structured data[edit]

integrating content from Commons, Wikidata, Wikisource and other projects into Wikipedia.

Note that the most recent RFC on using integrating Wikidata into English Wikipedia reached a severely split result. The discussion was extremely complicated, but I would oversimplyfy the result as follows:

  • Approximately 1/3 of the community was happy to integrate Wikidata.
  • Approximately 1/3 of the community wanted Wikidata integration eliminated.
  • Approximately 1/3 of the community were less savy on the technical issues involved, and they found both sides of the argument persuasive. They essentially said Wikidata integration would be acceptible if someone could wave a magic wand and we could get all of the positives and none of the negatives.

The Community is sorely stuck on this issue, and is very slowly trying to reach a consensus. Note: That RFC just about a year old. When I say the community is "slowly" trying to sort out the issue, I mean there has no visible progress in the last year and slim chance of a result in the coming year.

As a community member well familiar with this subject area, as a community member with extensive experience in the community consensus process, as a community member with a specialty in WMF-Community interaction, I offer you some advice. Any pressure from the WMF to promote Wikidata-integration-on-Wikipedia has approximately 100% chance of backfiring. Any percieved pressure from the WMF on this issue is likely to result in the existing partial Wikidata integration being stripped out by the community.

our content must become more structured, and we need to empower our contributors with tools to help them be effective and consistent in contributing and working with data in a more structured format.

It is unclear exactly what this would look like in practice. However please note the section above about Wikidata. The community consideres our free-form Wikipages to be our most important most valuable tool. There have been various conflicts in the past relating to the Foundation's desire to structurize things. If you build things that the community wants, all is fine. However words like "empower" and "help" are a fraud if you find yourself trying to shove unwanted things down our throat. You can't "empower" and "help" someone by waging war against them. Good intentions don't help, if you don't check whether people actually want to buy what you're selling.

25% of content consumed or created leverages “structured data”. This includes Wikidata but also extends to content from articles, templates, and other sources

Per the above points, this goal is at serious risk of failure. Any 25% target for structured data will fail horribly if EnWiki strips out existing Wikidata use, and more so if other major wikis reach a similar decision. Any pressure from the WMF on this issue would increase the risk of Wikidata use being stripped out, leaving few options for the WMF to actively pursue this goal.

Alsee (talk) 21:24, 16 May 2019 (UTC)

@Alsee: There has been some progress in the past year (e.g. adding citations to other databases for dates of birth and death), but I'd agree that it's nowhere near enough. The DBPedia grant is another step forward and could potentially enable the addition of millions of references.
Rather unfortunately, Wikidata is more than a decade younger than Wikipedia. In theory all of Wikipedia's infoboxes could be converted into "structured data", but until and if that becomes possible – and it's probably still years away – there probably won't be much further resolution on the matter. I would nevertheless agree with a goal of gradually increasing Wikidata usage, since in the meantime it's probably just going to mean external links and maybe infoboxes in some small subtopics. Jc86035 (talk) 10:48, 19 May 2019 (UTC)