Requests for comment/Wikimedia Enterprise

The following request for comments is closed. Comments were collected, nothing was proposed, and, as such, nothing passed.


This Request aims to collect comments and positions from Wikimedia movement stakeholders concerning the rise of the Wikimedia Enterprise project. The technical and philosophical aspect will be treated separately.

Background

The Wikimedia enterprise looks like a quite controversial project. Lot of information about the project available from its dedicated Meta-Wiki page, but nothing is reported on the Wikimedia Foundation Governance Wiki nor the Foundation official website until now.

However, the Wikimedia Enterprise Project is not an insignificant new project in the frame of Wikimedia ecosystem. It is the first commercial one in ".com" and it appears as a probable paradigm change of our initial and fundamental values.

Indeed, this project is in contradiction with different initial positions of the foundation :

  • Wikimedia Foundation statement of purpose : "The Foundation will make and keep useful information from its projects available on the Internet free of charge, in perpetuity."
  • Wikimedia Foundation Guiding Principles : "The Wikimedia Foundation is deeply rooted in the values of the free culture and free software movements. With the exception of "fair use" material, all information in Wikimedia projects can be freely shared, freely distributed, freely modified and freely used for any purpose, commercial or non-commercial, in perpetuity."

Technical question

Should we improve access to information from Wikimedia projects by improving the API system for commercial enterprises, as Wikimedia enterprise want to do?

I am not sure it is a fair description, as it is not about improving existing APIs, but building another one. --Base (talk) 21:42, 17 August 2021 (UTC)[reply]
Thanks fot this remark Base but building a new API is not "improving the API system" for you ? Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 22:25, 17 August 2021 (UTC)[reply]
Improving the API system includes building a new API venue for companies, that want/get special treatment for bribes. Grüße vom Sänger ♫(Reden) 07:24, 18 August 2021 (UTC)[reply]

Support

  1. Support Support A technical project increasing the diffusion of the Wikimedia project contain seams for me a good project, but I'm not competent enough to discussion the specificl IT issues, unfortunately. Lionel Scheepmans Contact 22:04, 16 August 2021 (UTC)[reply]
  2. Support Support Big corporations have big demands (e.g., around the performance and security of access). Providing a paid-for service to enable bespoke access in line with their requirements for these groups is acceptable. That is, making the information freely available to all is different to charging to provide custom technical access methods to that data that only benefit large corporates. QuiteUnusual (talk) 07:46, 17 August 2021 (UTC)[reply]
  3. Support Support This will reduce the load on our servers and allow us and the people that use our data more opportunities to get useful data in both directions. --Slashme (talk) 22:39, 17 August 2021 (UTC)[reply]
  4. I'm not sure many people will significantly oppose the need for a firehose API or more frequent tarball exports. I suspect all the discussion will be on how the cost of implementation and maintenance should be funded (if companies willing to pay do so for up-front costs or ongoing usage, pay at-cost or above-cost, or just be asked to make voluntary donations). The question of where raised money should be spent seems too broad for this RFC. T.Shafee(Evo﹠Evo)talk 05:04, 18 August 2021 (UTC)[reply]

Oppose

  1. ...

Other

  1. Doubtful It is doubtful There are two questions here. We should improve access to information from Wikimedia projects, if funds permit (and accounts show that they clearly do). I see no benefits to doing so with Wikimedia Enterprise. Certes (talk) 23:45, 16 August 2021 (UTC)[reply]
    @Certes, Wikimedia Enterprise is also a technical project with the target to enhance the API system. I'll change the question to make it clear. Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 23:52, 16 August 2021 (UTC)[reply]
  2. "It depends." First, if anyone has access to it, then everyone should have access to it. So if there is a high-speed API for Google (or whomever), then I want to be able to use it, too. Second, if there is an API built specifically for Google (or whomever) per their specs, then Google needs to pay for it, rather than using our donated funds. Levivich (talk) 15:45, 17 August 2021 (UTC)[reply]
    Google will pay for it; that's the point. 31.127.160.115 19:19, 17 August 2021 (UTC)[reply]
    No, not as I understand it. We will pay for its development using donated funds, in the hopes that we will later be able to sell a license to Google and others to recoup the costs. But that's not guaranteed. I'm talking about Google paying for its development, "cash up front." Levivich (talk) 22:03, 17 August 2021 (UTC)[reply]
  3. To put it very mildly, this is a Questionable RfC: Opened by a candidate to the Board of Trustees, two days before the election, and the first sentence of the RfC is a link to the voting instructions page (WHY?). Beyond that, the RfC does not look like well-thought for something of this scale. Sorry, I will not participate !voting here. MarioGom (talk) 19:05, 17 August 2021 (UTC)[reply]
    Hello MarioGom and sorry about this for this. I have already removed the sentence that you have a problem with. My intention was not to run an election campaign here, but to draw attention to the fact that this is an appropriate time to address such an important issue. It was probably awkward, but on the other hand, the whole election system pushes candidates to express their thoughts. I was asked to introduce myself, put up a picture, make a video, answer a bunch of questions, attend several meetings... And if you want to know what I think, I am against elections and for direct democracy. And I think that the call for comments is an excellent way to practice it. So feel free to give your opinion here and not vote for me in the elections. The calls for comments have more legitimacy to me than any opinion from the Wikimedia Foundation board. Sincerely, Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 22:18, 17 August 2021 (UTC)[reply]
    Global RfC openings are not a way for candidates to express their thoughts on campaign matters. This RfC inappropriately railroads the discussion with a bizarre framing of questions. Others have addressed this problem already in the comments below, so I'll avoid the redundancy. MarioGom (talk) 22:34, 17 August 2021 (UTC)[reply]
    Broadening a question scope ([1]), and changing it ([2]) after +25 people have already answered are a symptoms of the RfC not being ready to be open. MarioGom (talk) 22:51, 17 August 2021 (UTC)[reply]
  4. The claim "this project is in contradiction with different initial positions of the foundation" is false. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:23, 17 August 2021 (UTC)[reply]
  5. This is not a technical question. Harej (talk) 22:18, 17 August 2021 (UTC)[reply]

Philosophical question

Should we start selling access to information or more precisely "format in which data is delivered" from Wikimedia projects with Wikimedia enterprise?

And also in general situation (in reference to the10% of the gross total collected by the WMF for each book sold by Pediapress and so on)

Again, does not seem to be a fair description. Wikimedia Entrerprise does not sell access, but sells format in which data is delivered. As in it is about drivers for hire while one can still just grab a car for free and drive it themselves from our dealership of free cars of a sorts. Hope the simile works. --Base (talk) 21:46, 17 August 2021 (UTC)[reply]
Thanks again Base for this new technical consideration and enlightening analogy ! Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 22:32, 17 August 2021 (UTC)[reply]

Support

  1. Right now, a significant percentage of the money we donate to the Wikimedia Foundation goes to pay for Big Tech constantly sucking all our data down the internet tubes (for free). Why should we subsidize Big Tech with our donations rather than charging them a fair price for hogging most of our API usage? Framing this as "Wikipedia will no longer be free" is silly. It's more like "Wikipedia will no longer be freely exploited to enrich Big Tech". Nosferattus (talk) 00:09, 17 August 2021 (UTC)[reply]
    Silly or not Nosferattus, if we follow this point of view, we have to change the mission and vision of Wikimedia to adopt a business model. If it is the will of the movement, why not. But it will be without me in this case. Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 00:21, 17 August 2021 (UTC)[reply]
    "if we follow this point of view, we have to change the mission and vision of Wikimedia to adopt a business model" This sattemnt - for which you again offer no evidence - is false. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:45, 22 August 2021 (UTC)[reply]
    • For perspective, from the 2019/2020 financial report [3], the WMF spent ~2.4 million that year on internet hosting. Against over $129 million in revenue, this is 1.86% of the WMF's budget. Approximately as much as was spent on travel, over twice as much was spent on 'donation processing services', and the WMF's annual investment income was over twice the amount spent on internet hosting.
While 2.4 million is a lot of money, yes, calling it a 'substantial portion' of the WMF's budget is arguable. Jarnsax (talk) 16:16, 24 August 2021 (UTC)[reply]
  1. Overall all, I found their rationales essay quite well thought through. It's already possible to pay for wikimedia content in a printed format, with a portion going to WMF and a portion going to Pediapress, which isn't even a subsidiary (Sidenote: I'd like it to be easier to find the total income this brings to the WMF from this page). So long as the WMF clearly puts in place protections to ensure that free access to content will not be compromised i.e. licensing remains creative commons (CC BY-SA or better), database dumps and APIs remain in place and supported. So long as nothing is removed or restricted, then it seems ok to allow companies to pay for the bigger pipeline that is going to have to be built either way to avoid clogging up the free 'normal user' APIs. T.Shafee(Evo﹠Evo)talk 07:14, 17 August 2021 (UTC)[reply]
  1. Thanks for this comment @Evolution and evolvability. I'm not sure that Mediapress still work with the foundation regarding the fact that it's now impossible to use the book creator on Wikimedia project but what was sold by the company was the media not the text still on CC.BY.SA on the paper. And thanks to point that Wikimedia Foundation "receives 10% of the gross total for each book sold". This will broaden the debate.
    For the question of licence, don't you know that Wikidata and a part of Commons are already on CC0 ? Big tech already to use all this contains without any obligation ? Which means that Wikimedia Foundation has already stopped to "clearly puts in place protections to ensure that free access to content will not be compromised" as you wrote. With the venue of Wikimedia enterprise and Wikimedia abstract on wikidata (in CC0) the Wikimedia movement Will start to be the Big fabric of Big Tech information and Wikimedia Foundation the middleman, not to say the slaver. Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 12:44, 17 August 2021 (UTC)[reply]
    Apologies, I was unclear about licenses when I said "CC BY-SA or better"; I should have said CC BY-SA or more permissive/open (I'd actually prefer CC-BY as default for text & images to be honest, but we're stuck with -SA). The main risks to avoid would be: A) avoid more restrictive -NC or -ND variants, and certainly not locking any content as all-rights-reserved, B) throttling of free APIs etc in order to drive users to a paid pipeline. However I think that explicit preventative policies can be put in place for both of these. T.Shafee(Evo﹠Evo)talk 04:46, 18 August 2021 (UTC)[reply]
  2. I'm in both camps. Selling access to the data - no we shouldn't. Selling a technical service that is customised to allow a big corporation to access the data in a way that benefits that corporation - yes we should. The corporation can then choose to use the generic free service, and work around it, or pay the WMF to provide something bespoke to their needs. This isn't a deviation from the core mission of the Foundation. QuiteUnusual (talk) 07:48, 17 August 2021 (UTC)[reply]
    Thanks for this point of view QuiteUnusual. On my side, I see the deviation at the level of selling not the service for which I am for but in a fair way for all. Selling a privileged access service is to favor the richest over the poorest. Or in other words making the rich even richer and the poor even poorer since they will have to give their personal data to benefit from the free service of big tech. Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 13:32, 17 August 2021 (UTC)[reply]
  3. Useless and biased RfC, but I kind of have to participate given the chorus above. WMF isn’t going to be ‘selling access to information’, whatever that is, WMF is going to be (essentially) ‘selling service-level agreements to re-users of Wikimedia content’. The only possible problem with that is whether that will actually return a profit or not, which is practical, not philosophical. It is certainly not the worst use of money for WMF (costly events loved by Wikimedians come to mind first). [@Organiser, please do not respond to my comment like to all the others if possible.] stjn[ru] 18:52, 17 August 2021 (UTC)[reply]
  4. Our content is distributed for free, and for-profit Big Tech companies use it to make money. If they want 'enterprise features' offered by the WMF, it seems perfectly reasonable to me for the WMF to charge them access for it. Those same funds can perhaps go into Foundation activities, as well as developer recruitment to start implementing more features and fixing more phab bugs, etc. I sympathise with Levivich's oppose concerns regarding fund allocation. However, in my eyes that is separate from whether this initiative is inherently bad, which is what this poll seems to be about. ProcrastinatingReader (talk) 18:57, 17 August 2021 (UTC)[reply]
  5. It's like with Linux distributions with support services. The system underneath is naturally under a free license, and will stay like that (that's the purpose of copyleft licenses, right). What you pay for is guaranteed support -- you will not be left alone figuring out what to do. That's helpful to big companies -- often, paying for a professional guaranteed service is easier (and in result, cheaper) than doing it all themselves. Something very similar applies for Wikimedia Enterprise. There are big tech companies that make use of the free knowledge base we all help to maintain. Now, they make their own systems to do that thing, with Wikimedia community having exactly 0% of the profit they make thanks to that back. With Wikimedia Enterprise, a WMF subsidiary will create such a system for them, guaranteeing its stability and usability. Wikimedia community will get (hopefully -- but that's a practical question) something' out of it through the Foundation, which is a good thing. No one is trying to sell the data. It is (and always will be) possible to download raw data and play with it. --Martin Urbanec (talk) 19:00, 17 August 2021 (UTC)[reply]
  6. Bonjour. Je vois dans Wikimedia Enterprise plusieurs avantages intéressants :
    • Poser un cadre officiel sur des récupérations massives d'informations par des grandes entreprises. Cela permet de différencier clairement ce qui est autorisé de ce qui ne l'est pas, car l'abus de droit est un concept mal connu, surtout à la Silicon Valley.
    • Diversifier les sources de revenus du Mouvement Wikimedia est une manière de pérenniser le mouvement d'un point de vue purement financier.
    • L'accès à l'information n'est aucunement diminué pour tous les autres lecteurs ou utilisateurs. Et plus encore, faire payer cette information à certaines parties prenantes permet de faire un juste rappel que non, tout n'est pas gratuit en ce bas monde. --Consulnico (talk) 15:49, 18 August 2021 (UTC)[reply]

Oppose

  1. Oppose Oppose Starting to sell accessibility to Wikimedia project contain is contrary to the mission and vision of the Wikimedia Movement and Foundation. The idea of increasing the capacity of the movement by selling access to raise more money is not smart. Going to the feel of Web of the Big Tech will drastically decrease the power and the independence of the movement. The money will never be a problem for Big Tech company, but always a solution to increase their power and influence on our movement with commercial contracts and all the legal pressure linked to them. As sharing and non-profit movement, Wikimedia should not be interested in the fact that commercial companies make money with Wikimedia project contain, especially since we don't have difficulty to collect our own money from donators. Our power is our soul. A non-profit, inclusive, democratic, sustainable and generous movement that inspires the respect of the outside world. Selling the access to our information is selling our soul to the commodification. Lionel Scheepmans Contact 22:04, 16 August 2021 (UTC)[reply]
    @Lionel Scheepmans: We're already heavily commodified by Big Tech. The question is, should we start charging for this commodification. (See my comments in the support section above.)Nosferattus (talk) 00:12, 17 August 2021 (UTC)[reply]
    @Nosferattus: for me the philosophical question is : do we want participate to the commodification by playing the game of big tech or not  ?
  2. Oppose Oppose per above. SHB2000 (talk | contibs) 23:21, 16 August 2021 (UTC)[reply]
  3. Oppose Oppose Once editors realise that they have become unpaid interns of a for-profit, they will walk away, and our remarkable achievements will rot and die. Certes (talk) 23:39, 16 August 2021 (UTC)[reply]
  4. Oppose Oppose There are things we can & should do, like aggressive enforcement of the attribution clause in licenses, or asking big tech firms that use our stuff for donations & publicly mocking any that do not donate. This is not one of them. Pashley (talk) 01:53, 17 August 2021 (UTC)[reply]
  5. Oppose Oppose Our goal is to make knowledge free, so it doesn't matter if companies profit from it. Respecting our principle that we will not gain from knowledge and will not deceive volunteers is more important than a few dollars. Once we take money for our services and content and trade knowledge for money, we are left with the question of when will Wikipedia become a for-profit organization? And have we been serving over the past years a for-profit organization for free? Arabic native speaker; please excuse any errors on my part--جار الله (talk) 02:23, 17 August 2021 (UTC)[reply]
  6. Oppose Oppose I didn't like editing on Wikitravel, because the work I volunteered was being used by the corporation that ran it for profit, and they subsequently ran it into the ground, causing the split. I do not countenance Wikimedia going down that road. Ikan Kekek (talk) 02:31, 17 August 2021 (UTC)[reply]
  7. Oppose Oppose per Lionel Scheepmans. --Holder (talk) 03:47, 17 August 2021 (UTC)[reply]
  8. Oppose Oppose WMF does'nt have to sell what we all do for free. --Croquemort Nestor (talk) 05:54, 17 August 2021 (UTC)[reply]
  9. Oppose Oppose per Lionel. Mathis B (talk) 06:34, 17 August 2021 (UTC)[reply]
  10. Oppose Oppose per above.--Jean-Mahmood (talk) 06:38, 17 August 2021 (UTC)[reply]
  11. Oppose Oppose as Lionel, and I fail to see any need for even more money, we can't spend the vast amount we already have useful for our mission, why sell our soul for even more unneeded money? Grüße vom Sänger ♫(Reden) 08:37, 17 August 2021 (UTC)[reply]
  12. Strong oppose - strongest possible for every reason that's been given here. It's already a scandal that the volunteers' contributions to articles and their work to keep the project clean raise the funds to pay a bloated employee force of nearly 500, some of whom are on well over 6 figure salaries plus perks. Not only is it totally contrary to the founding principle, but the WMF is already so over-funded that it's lining its pockets with an endowment, and constantly creating paid positions to manage its offshoots, and constantly thinking up more time-wasting schemes to pay more staff. There would probably be a massive walkout if this ever came to pass. Kudpung (talk) 09:26, 17 August 2021 (UTC)[reply]
  13. Charging Big Tech for their high-volume usage isn't necessarily a bad thing in and of itself, but not via Wikimedia Enterprise LLC as it is currently proposed because the necessary controls aren't in place to make sure donated funds are properly used. (They're not in place at WMF, Inc. or the Wikimedia Endowment either.) Levivich (talk) 15:51, 17 August 2021 (UTC)[reply]
  14. Oppose Oppose. in agreement with Lionel----Fuucx (talk) 16:12, 17 August 2021 (UTC)[reply]
  15. Oppose Oppose I would only support if the community has a 100% say in how the money is spend. Natuur12 (talk) 16:39, 17 August 2021 (UTC)[reply]
  16. Strong oppose שי אבידן (talk) 17:28, 17 August 2021 (UTC)[reply]
  17. Oppose Oppose Obviously, no, the Foundation should not sell access to "information from Wikimedia projects." This is, however, irrelevant to Wikipedia Enterprise, which does not "sell access to information." The information is publicly accessible, and will remain so. The Foundation is selling none of that. No closure on this question by stewards will change anything about the future of the Wikimedia Enterprise project. Zoozaz1 (talk) 20:20, 17 August 2021 (UTC)[reply]
  18. Oppose Oppose per above. --Assyrtiko (talk) 17:28, 19 August 2021 (UTC)[reply]
  19. Oppose Oppose --Jan Kameníček (talk) 08:01, 22 August 2021 (UTC)[reply]
  20. Oppose Oppose worrisome. 4nn1l2 (talk) 10:13, 3 December 2021 (UTC)[reply]

Other

  1. yes and no. I don't think WMF should be paying to handle large volumes of requests for other organizations, commercial or not, effectively removing these resources from direct users of WMF projects. Jura1 (talk) 11:13, 17 August 2021 (UTC)[reply]
    @Jura1 That's effectively happening now though -- the only way how other orgs can reuse data is through the publicly available APIs (and/or dumps). Moving such usage to a dedicated paid service will free resources on the main cluster, as well as increase revenue. Martin Urbanec (talk) 19:02, 17 August 2021 (UTC)[reply]
  2. This question is framed preposterously. We don't hold RFCs around vague philosophical ideas. Harej (talk) 20:33, 17 August 2021 (UTC)[reply]
  3. This isn't at all what is being proposed. The idea isn't about selling access to our data. It's about recognising that large organisations that re-use our data are scraping it at the moment anyway, and that can get more from it if we provide it in a suitable form, we'll have lower overheads, and they're willing to pay for a better service. --Slashme (talk) 22:42, 17 August 2021 (UTC)[reply]
    We don't know if they're willing to pay for it, they've made no commitments to do so. Not sure what you mean by lower overheads, but this will cost us donated funds to implement, and it won't reduce our expenses, it'll actually increase them as we'll have to pay to maintain it. They're scraping it at the moment and this will provide a dedicated feed to them, but at our expense, with no guarantee about whether anyone will use it or pay for it or how much they'll pay. Levivich (talk) 16:08, 18 August 2021 (UTC)[reply]
Without wishing to impose upon the discussions on this page, I do want to add in direct reply to Levivich's comments here in case it is helpful for them or others. 1. There have been comprehensive and extensive 'discovery' product meetings undertaken with many potential customers of this service. The team is very aware of the product requirements and desires of many companies, and building a service that will satisfy as many of them as possible as quickly as possible is the goal of the "version 1" product deliverable. The product roadmap (which is being refined as we go, and documented on our pages on Phabricator and mediawiki.org) will add more of these future-customers' desired features. 2. The fact that you have not seen these discussions with these companies to ascertain their willingness to commit to this service, and the fact that we have not made public announcements of the existence of discussions with each potential customer, does not mean that those discussions are not happening. 3. Our hope is to be able to make the announcement of the first confirmed customers alongside the launch of the product website itself in a couple of months' time. Further customers will, we expect, be able to be announced before the end of the year. 4. While there is a setup cost (we have to build the product before they will pay to use it, like anything) the service is not going to be run 'at a loss' - it will generate revenue for the wikimedia movement. The expenses of the service will be more than covered by its revenue, it is a business, after all. If 'Wikimedia Enterprise' is unable to do that in the long term, it will be closed. I hope that clarifies matters. LWyatt (WMF) (talk) 18:05, 23 August 2021 (UTC)[reply]
it is a business, after all launched using money donated to a non-profit. Another one of those "not necessarily illegal, still a bad idea" WMF ideas.
And neither you nor anyone else can guarantee that revenue will exceed expenses. Over 90% of new businesses fail, after all. If this business doesn't generate revenue in excess of expenses, then the losses incurred means donated funds wasted. And you didn't mention how much money WMF has already spent and plans to continue spending on this; over $1 million spent and over $1 million to go, I'm guessing. So millions of dollars of donated funds at risk here, amirite?
And you don't have to answer this but I'm going to take another guess here, because I don't believe that there is nobody at WMF who has heard of the concept of pre-selling custom software. So I'm guessing the reason this software isn't pre-sold (no deposits paid, no contracts signed) is because during those many discussions, WMF asked the potential customers about pre-sale, and the customers declined. The customers said build it to our specs first and then we'll decide if we want to pay for it, and WMF decided to go ahead and spend donated money on doing just that. I know those discussions are confidential and WMF won't confirm or deny if they offered pre-sale, but that's my guess.
One of the reasons I worry that the customers won't end up paying for this and a bunch of donated funds will be lost is that I don't have confidence that WMF will be able to build software that satisfies the customers' expectations. My belief is admittedly biased as I am a user of WMF software. Levivich (talk) 19:19, 23 August 2021 (UTC)[reply]
Dear @Levivich,
While, I can't give a precise timeline, but we have committed to: preparing a public cost-accounting of the 'startup costs' of the project itself; and also that the project's revenues will be declared in the context of the WMF's audited financial reports [the "Form 990"], or equivalent. This will make it possible for anyone to compare - so that you don't have to take my word for it - that the project is more than covering its setup and ongoing costs.
On a personal note, I find that most people are concerned that the project would raise too much in revenue and thereby distract from the movement's non-profit mission. This is why we are instituting the policy that Enterprise would never be more than 30% of the WMF's revenue in any given year. However, there is the equal-and-opposite concern that you are expressing: that the project would raise too little and become a net-loss for the movement. As stated above, we should be announcing the first tranche of customers in a couple of months, which should allay that latter concern.
Finally, as for 'pre-sold' software: There is a great diversity in the product requirements that different potential customers are looking for. And there is also a great diversity in the ways that these companies currently access Wikimedia sites' content. It is not a 'build it once and ship that' product. You can read about the background research and product roadmap on the MediaWiki page "overview" section here: mw:Wikimedia_Enterprise#Overview.
I hope this extra information helps.
As I do not wish to overwhelm/flood this page with details, I encourage any further questions/comments should be posted on Talk:Wikimedia Enterprise. Sincerely, LWyatt (WMF) (talk) 17:09, 24 August 2021 (UTC)[reply]
  1. # Pas opposé Mitigé/Neutre (pour une réflexion auparavant). Ce projet n'empêche pas l'accès libre et gratuit à l'information, provenant des différents mouvements wikimedia. Je suis pour qu'il y ait suffisament de garde-fou (que ces API soient accesible à tout organisme non lucratif, seuls les grosses entreprises qui se font de l'argent sur notre dos devront payer). Lionel Scheepmans, tu parles de dépendance aux GAFAM, mais c'est déjà le cas actuellement, via les dons. Oui, notre mouvement à besoin de ressources financières en suffisance. J'ai parfois l'impression qu'on en manque, d'un point de vue tech, quand je vois le manque de réactivité aux requêtes Phabricator, et à la mise en place de nouvelles fonctionalités sur les différents wikis, et aussi d'un pdv ressource humaines quand je vois la difficulté de faire le liens avec les wikis (messages non traduits, pas de réponse aux commentaires...). WMF compte 400 employés, ce qui n'est pas énorme comparé à d'autres organisanisations internationales. Même si bien sûr, le principe même de notre mouvement reste le bénévolat, nous avons besoins de salariés pour tous ce qui a été dit plus haut, et de revenus stables pour mener à bien nos différents projets dans le futurs et ainsi s'assurer de la pérénité du mouvement ! -- Nemo Discuter 07:54, 19 August 2021 (UTC)[reply]
    (in English below)
    Bon, j'avais dit dans la page de discussion que je ne reviendrais plus sur cet appel à commentaire avant la fin des élections puisque cela dérange, mais si on me notifie, ma politesse m'oblige à répondre.
    Nemo|Tu parles de "réflexion auparavant" et c'est déjà un point manquant dans ce projet sur lequel travail déjà près d'une dizaine de personnes sans qu'une grande part de la communauté d'éditeurs ne soit au courant. Quand aux API elles existent déjà et sont gratuites, on peut les améliorer pour tous et ce serait une bonne chose. Mais ce qui se pointe à l'horizon ressemble plus à une autoroute payante à côté de la route nationale.
    Tu peux ajouter une centaine d'employés à ton estimation et il n'y a pas la moitié qui code sur nos projets. Les détails se trouve ici. Si l'on doit étendre le conseil d'administration pour surveiller tout ça, c'est que c'est déjà de trop. Ajouté à cela, plus il y aura de salarié, moins il y aura de bénévoles. Tu te vois contribuer bénévolement à côté de personnes rémunérées toi ? Et ben dis-toi que sur phabricator, ça doit être pareil. Et voilà maintenant que l'on parle de rémunérer les administrateurs dans les projets...
    L'entreprise qui se fait le plus d'argent sur notre dos est la fondation Wikimédia. Ce n'est pas une entreprise, tu me diras. Ok. Et Wikimédia Enterprise .com, toujours pas ? Ils veulent plafonner à 30 % des revenus totaux. Ok. Ben si c'est comme ça, il faudra que Wikimedia Enterprise gagne toujours plus d'argent pour que l'ensemble de la fondation en gagne aussi toujours plus. Un deuxième cercles vicieux.
    Et la chose le plus importante qu'il faut comprendre. Quand la Fondation fait du partage un don, c'est une première perversion. Nous les contributeurs, nous partageons notre temps ensemble pour partager la connaissance sans rien attendre en retour, mais la fondation a fait de notre travail un don auquel les lecteurs doivent maintenant donner un contre-dons en argent et pas en temps. D'où selon moi, la diminution de la participation dès l'arrivée des campagnes au sein des projets.
    Sauf que là, on en arrive à un nouveau stade de perversion. Il est plus question de donner, mais de vendre. Et tu sais la différence entre les deux ? C'est le contrat. Le contrat de don n'existe pas, sauf peut-être moralement, mais toujours sans obligation légale. Par contre, si la fondation commence à vendre, ben elle va aussi devoir signer des contrats avec des entreprises à qui je ne prêterais pas ma chemise et certainement pas mes données à caractères personnelles. Regarde un peu leur loyauté...
    Les dons des Gafam n'entraine aucune dépendance. La fondation ne doit rien en retour d'un don et c'est d'ailleurs là l'avantage du statut de fondation. Par contre, si elle commence à vendre, c'est plus pareil. Elle est tenue de négocier la vente, d'assurer le service, de répondre aux plaintes, aux attaques en justice, etc. Et ça, c'est un nouveau cercle vicieux, car il faudra engager plus de personnel pour négocier les contrats, assurer les services, répondre aux plaintes et aux attaques en justice.
    Voici donc mon analyse à ce stade de compréhention. Bien à toi, Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 00:29, 20 August 2021 (UTC)[reply]
    ----
    Well, I had said in the discussion page that I would not come back to this call for comments before the end of the elections since it is disturbing, but if I am notified, my politeness obliges me to answer.
    You talk about "thinking before" and this is already a missing point in this project on which about ten people are already working without a large part of the editor community being aware of it. As for the APIs, they already exist and are free, they can be improved for everyone and this would be a good thing. But what is on the horizon looks more like a paying highway next to the main road.
    You can add a hundred employees to your estimate and not half of them code on our projects. details are here. If we have to expand the board of directors to oversee all this, it's already too much. Added to that, the more employees there are, the fewer volunteers there will be. Can you see yourself contributing voluntarily next to paid people? Well, you can imagine that on phabricator, it must be the same. And now we're talking about paying the administrators in the projects...
    The company that makes the most money on our backs is the Wikimedia Foundation. It's not a company, you'll tell me. Okay. And Wikimedia Enterprise .com, still no? They want to cap it at 30% of total revenue. That's huge! And you know what else? If it's ever like that, then Wikimedia Enterprise will have to make more and more money so that the whole foundation will make more and more money. A second vicious circle.
    And the most important thing to understand. When the Foundation makes sharing a gift, it is a first perversion. We contributors share our time together to share knowledge without expecting anything in return, but the foundation has made our work a gift to which the readers must now give a counter-gift in money and not in time. Hence, in my opinion, the decrease in participation as soon as the campaigns arrive within the projects.Except that now we have reached a new stage of perversion. It's no longer about giving, but about selling. And you know the difference between the two? It's the contract. There is no such thing as a donation contract, except perhaps morally, but still without legal obligation. On the other hand, if the foundation starts selling, well, it will also have to sign contracts with companies to whom I wouldn't lend my shirt and certainly not my personal data. Take a look at their loyalty...
    Gafam's donations do not lead to any dependence. The foundation doesn't owe anything in return for a donation and that's the advantage of being a foundation. On the other hand, if it starts to sell, it is not the same. It has to negotiate the sale, provide service, respond to complaints, legal action, etc. And that's a new challenge. And that's a new vicious circle, because you have to hire more staff to negotiate contracts, provide service, respond to complaints and lawsuits.
    Here is my analysis at this stage of understanding. Lionel Scheepmans Contact French native speaker, sorry for my dysorthography 00:51, 20 August 2021 (UTC)[reply]

Discussion concerning technical issue

Discussion concerning philosophical issue

  • I believe that the most rational and compromising thing to do for the Foundation would be to ask for a fee to corporations, and provide free access to open source projects and non profit users/organisations. --ValeJappo【〒】 08:54, 17 August 2021 (UTC)[reply]
    @ValeJappo Well, that's (more or less) happening. Orgs which need real time access point with guaranteed stability will have to pay a fee (there's really no imaginable way to guarantee stuff that runs for free), the rest of the users will be able to either use the current APIs/dumps, or download regular dumps of the enterprise data. Martin Urbanec (talk) 19:04, 17 August 2021 (UTC)[reply]
  • Concurring with Justin: The WMF already gets big donations from Big Tech. The main issues are that even if the WMF thinks it has found another source of additional revenue - and there is absolutely no shortage of money right now - what are they going to do with it and how would the volunteers who provide the content for free benefit from it? Better software? A better Wikimania? History has demonstrated that nothing much comes the volunters' way, so why does this need to be done at all?Kudpung (talk) 05:09, 18 August 2021 (UTC)[reply]

Comment from Wikimedia Enterprise community liaison

Dear all,

I've created a new subheading in order to deliberately separate this comment, as a WMF employee working directly in the Enterprise team, from the wider RFC – to ensure it does not appear as if I am interceding in the free-flow of community processes.

  • Above all else, I wish to clarify something that many people are concerned about in their comments above: The Wikimedia Enterprise API suite is not “selling” Wikipedia. The existing tools that anyone can use to copy Wikimedia content remain the same and in place, and anyone [including commercial orgs] can still use them. Obviously the copyright license(s) of the contents of Wikimedia projects remains the same too - that cannot be altered. To draw an analogy: this is not selling water, it is creating a new pipe that businesses with specific demands for a reliable flow of water can purchase access to. In order to contractually guarantee the reliability of the flow of that water we must build especially robust pipes, and have dedicated pumps (and staff to maintain those pumps). Those requirements are very specific, but currently those companies just connect their factories to the general water supply for the public. That status quo is neither good for the needs of those companies, nor for their customers, nor for the general public using the normal pipes, nor for the public pipes infrastructure, nor for the staff maintaining those public pipes. Instead, we are building a special system that they can pay for to access, if they prefer it over the public system, designed for their specific needs. This is in accordance with the Movement strategy in two ways: to improve the [financial] sustainability of the movement, and to improve the user experience of our readers who access our content 'downstream' (not via our websites directly). Also, as the Essay discusses at length - the provision of commercial services is entirely consistent with a F/LOSS ecosystem. The Enterprise project allows for the Wikimedia movement to diversify (not maximise) the revenue that it receives, by building services for the companies that are already commercialising our content for their own purposes.
  • As the RFC's "background" section notes, the talkpage of the main Meta page for the project has had a lot of conversations.The talkpage itself has had ~50 different editors and 80+ watchers. The concerns raised in the "oppose" !votes above have been talked through by people on that page, as well as during the various video "monthly office hours" (videos linked on the project mainpage) and at SWAN, Clinic, EMWCon, Wikimania sessions. As far as my memory serves, I do not recall seeing any of the usernames of the above !oppose votes appearing in any of those previous fora over the last 5 months (since the extensive documentation was published and translated for community comment and feedback, in March). The topic has also been discussed extensively before - most notably twice in the Strategy itself. I take this fact as a sign that I did not sufficiently advertise the existence of the project's Meta talkpage, and the office hours meeting, and contacting me directly, as methods by which people can raise their concerns and questions.
  • With regards to some of the specific concerns raised – e.g. of financial transparency and community control of the eventual revenue – I encourage you to look through our Operating Principles. These include things such as: that Enterprise revenue will never exceed 30% of WMF total revenue; that no customer will have access to 'exclusive' contracts; that the eventual content provided via these APIs is the same as that which is available via public services, etc. etc.
  • Finally, for the technically minded – free (both 'gratis' and 'libre') access is available to anyone, without contracts, to the 'daily dump + hourly diff' Enterprise API via Wikimedia Cloud services already. You can use it today. The 'two week dump' will be available soon via the normal dumps page.

If you have any specific questions, please feel free to contact me directly - or to leave a message on the project's talkpage. Equally, if you have a community group (e.g. affiliate) who would like to have a meeting with the WIkimedia Enterprise team to discuss any of these issues, we are at your disposal. Please contact me and we can arrange a time, software platform, and preferred language for such a dedicated discussion. Sincerely, LWyatt (WMF) (talk) 13:14, 18 August 2021 (UTC)[reply]