Wikipedia:Village pump (WMF)

 Policy Technical Proposals Idea lab WMF Miscellaneous 
The WMF section of the village pump is a community-managed page. Editors or Wikimedia Foundation staff may post and discuss information, proposals, feedback requests, or other matters of significance to both the community and the Foundation. It is intended to aid communication, understanding, and coordination between the community and the foundation, though Wikimedia Foundation currently does not consider this page to be a communication venue.

Threads may be automatically archived after 14 days of inactivity.

Behaviour on this page: This page is for engaging with and discussing the Wikimedia Foundation. Editors commenting here are required to act with appropriate decorum. While grievances, complaints, or criticism of the foundation are frequently posted here, you are expected to present them without being rude or hostile. Comments that are uncivil may be removed without warning. Personal attacks against other users, including employees of the Wikimedia Foundation, will be met with sanctions.

« Archives, 1, 2, 3, 4, 5, 6, 7

No relicensing template edit

How is Template {{WikimediaNoLicensing}} permissible under wmf:Term of Use? It's even fully protected. User Anthony (inactive) created the original in 2004, and the license only dates to 2011, so maybe the template is grandfathered in somehow? Beyond the legal, I've seen it on a user page (and elsewhere) and on user pages it seems to be a declaration that the user does not intend to comply with the Terms of use, and even if the statement above the Publish button negates that from a legal point of view (does it?) which maybe means the template's assertion is void, it hardly seems the right attitude for a User here to have. Should it be taken to Afd? Even if legally void, why encourage that with a template, even if it's just a pointless sign of an ornery user strutting some attitude on their user page, like a lot of userboxes are. Adding Slaporte (WMF). Mathglot (talk) 19:02, 4 March 2024 (UTC)Reply

It looks like there about eight users active in the past year who display that template on their user page. Most users displaying the template have not edited in 10 years or more. I don't see how displaying that template can override the terms of use. I don't see it as a major problem. If the template does not have any legal effect, then trying to delete it may create more drama than it is worth. Donald Albury 20:43, 4 March 2024 (UTC)Reply
  You are invited to join the discussion at Wikipedia:Templates for discussion/Log/2024 March 4 § Template:WikimediaNoLicensing. — Frostly (talk) 21:12, 4 March 2024 (UTC)Reply

Al-Quds University edit

Why is Al-Quds University not only mirroring the English Wikipedia—which I presume is permitted by law—but also using Wikimedia logos? TrangaBellam (talk) 15:13, 9 March 2024 (UTC)Reply

@Slaporte (WMF): - FYI. TrangaBellam (talk) 15:17, 9 March 2024 (UTC)Reply

Conversations with the Trustees - next call this Thursday 21st! edit

Hi all. I just wanted to give you a heads-up, in case you didn’t already know, that there are regular ‘Conversation with the Trustees’ events that you are welcome to attend and ask questions to the Wikimedia Foundation Board of Trustees (who oversee and guide the Foundation). I’m hosting the next one, taking place this Thursday 21st March at 19h UTC and, speaking as a long-time enwiki editor, it would be really nice to see people from here attending and engaging in the discussions. Please see m:Wikimedia Foundation Community Affairs Committee/2024-03-21 Conversation with Trustees for details! Thanks. Mike Peel (talk) 21:51, 18 March 2024 (UTC)Reply

This is good to know about, Mike; thanks for sharing! Sdkbtalk 18:31, 19 March 2024 (UTC)Reply

Proposal: WMF should hire a full-time developer to do basic maintenance on MediaWiki edit

I've been disappointed with the state of disrepair of MediaWiki for years, but yesterday I've become aware of an issue that finally drove me to complain: there was a basic SVG rendering bug that has been fixed upstream 4 years ago, but it still torments Wikipedia readers because WMF can't be bothered to install the fixed version T97233. WMF also refuses to switch to a less buggy SVG rendering library T40010 or to let the browsers do the rendering themselves T5593. Other users there expressed skepticism that SVGs would ever work here and we should revert to PNGs instead, as such issues have existed for more than a decade without being addressed.

This lack of basic maintenance is not limited to SVGs. There is also the well-known issue that graphs are "temporarily" disabled, which was triggered by MediaWiki using the Vega 2 library for 6 years after its end-of-life, until this time bomb exploded in their face. It looks like the current "solution" is just disable graphs forever T334940.

Another issue is that MediaWiki still runs on Debian Buster, the Debian stable from two releases ago. Fun fact, it will be end-of-lifed in three months, so we'll have one of the biggest websites in the world running on unsupported software. And these are only the problems I have personally encountered. Other editors tell of many more that I won't list here.

One might think that this situation is due to a lack of funds, but this is not the case. WMF has so much money that it doesn't know what to do with it: Signpost May 2023, Signpost August 2023. That's why I'm launching this proposal to tell it: hire a full-time developer to do at least basic maintenance. It's unconscionable to donate millions of dollars to other charities while your own website is falling apart.

It would be in fact perfectly natural natural for such a wealthy foundation administering such a large website to fix bugs themselves, or even take over development of the libraries it depends upon. I'm not demanding that much. Only to keep the software stack remotely up to date. Right now it's downright archaeological. Our billions of readers are suffering through issues that the rest of the world has long solved. Tercer (talk) 15:56, 23 March 2024 (UTC)Reply

As I understand it, the WMF has hundreds of staff and these include developers. Github has 558 names of such. So, my impression is that there's no lack of staff or other resources. Presumably it's more matter of organisation and fit. I'm guessing that there's a lot of legacy code and technical debt and maybe this is too brittle and rotten to maintain easily. The graph debacle indicates that senior management ought to be getting a grip on this before a more catastrophic gap opens up. Andrew🐉(talk) 17:58, 23 March 2024 (UTC)Reply
Obviously WMF has some developers. Certainly not hundreds, let alone 558. In any case none of them is dedicated to maintenance, otherwise Wikipedia's servers wouldn't be in a worse state than my grandmother's PC. I assume they are working on sexy new features such as the visual editor, the reply function, or the dark mode. Maintenance is boring, and doesn't look impressive in your CV. Nobody wants to do it. That's why I'm proposing a full-time maintainer.
Your alternative theory that they have enough resources but still can't do maintenance can be summarized as rank incompetence, and that's too cynical for my taste. It's also not actionable. What could one propose? "Proposal: WMF should get its shit together"? Tercer (talk) 19:09, 23 March 2024 (UTC)Reply
The WMF does appear to have hundreds of developers and engineers. For example, see Developers/Maintainers which has a specific column documenting maintenance responsibilities. Some of these are the responsibility of entire teams such as Wikimedia Site Reliability Engineering which has a headcount of about 45. There are still clearly gaps in this structure, as shown by the year-long outage of graphs, for example. But the idea that there's nobody currently responsible for maintenance in a general sense seems too simplistic. The problem seems more that there's a complex structure in which it's easy for issues to fall down cracks or for people to pass the buck. Andrew🐉(talk) 21:21, 23 March 2024 (UTC)Reply
I took a look at the gigantic list in Developers/Maintainers; the first two names there are volunteers, not staff, so we can easily discount that as indicating that WMF has hundreds of devs. All the ones I clicked in Wikimedia Site Reliability Engineering, however, are actually staff, so we can take 45 as a lower bound for the number of devs. Fair enough, some of them are responsible for maintenance, but it's clearly not enough. The WMF can easily afford to hire another, and it should urgently do so. Tercer (talk) 23:07, 23 March 2024 (UTC)Reply
I find these threads tiresome. By background, I'm a real software engineer for real money in real life. I do some software development on enwiki-related projects as a volunteer. The pay sucks (but it's no worse than I get paid for editing) but at least I get to pick and choose what I work on, when I want to work on it, and how I want to architect it.
I've found my interactions with the WMF development staff similar to my interactions with any dev group I've ever worked with. Some are good, some not so good. There's a few who are absolute joys to work with. There's a few who are grumps. But then again, you could cross out "WMF developer" and write in (with crayon if you like) "enwiki editor" and you would still have a true statement.
The basic architecture is 25-ish years old. There's a lot of legacy crud. The fact that the core system is written in PHP just boggles my mind. Recruiting must be a challenge. How do you attract top-shelf talent when what you're offering is an opportunity to work on a legacy code base written in PHP and salaries which I can only assume are not competitive with what the Googles and Facebooks and Apples of the tech world are offering. And yes, you are right, doing maintenance work is not what people want to do. If you told somebody, "Your job is to ONLY work on maintaining the old stuff and you'll never get a chance to work on anything that's new and shiny and exciting", I can't imagine you'd get many applications. RoySmith (talk) 23:54, 23 March 2024 (UTC)Reply
And the slow code review process discourages the volunteers who are affected by longstanding bugs from working on fixing them. And of course a company with a known-bad workplace culture.
I think there are people, myself included, who would be willing to work on only fixing bugs rather than building new things in principle, but probably a lot of those people (again including myself) have internally vilified the WMF for exactly this reason so would consider it to morally repugnant to work for them. * Pppery * it has begun... 00:32, 24 March 2024 (UTC)Reply
I am one of those people. The experience described in that link is totally unacceptable and would lead to prosecution in many jurisdictions. I am ashamed to be associated with its perpetrators. Certes (talk) 01:41, 24 March 2024 (UTC)Reply
That's why WMF has to pay them. You'll never get boring infrastructure work done by volunteers. And no, I don't believe you'd have any difficulty finding applicants if you offer a decent salary. Even for working with PHP (it's no COBOL, everybody knows PHP). WMF can afford to pay even a top salary from a tiny fraction of the money it has been setting on fire. Tercer (talk) 10:26, 24 March 2024 (UTC)Reply
Yes, that sounds much more likely to succeed than giving the interesting work to the paid staff and hoping some mug will volunteer to do the grind for free. One option is make dedicated maintenance a role rather than a person, and to allocate it to a different member of staff each month. (Other time periods are available.) That way no one has to do it for long enough to drive them to resignation, and it's a chance to cycle the skill set with e.g. graph maintenance being done when a graph expert is on duty. Certes (talk) 11:44, 24 March 2024 (UTC)Reply
Small bug fixes can be spread out amongst developers. But truly addressing significant technical debt means cleaning up the software framework to be more sustainable. That's not something that can be done effectively by rotating the work. isaacl (talk) 20:49, 25 March 2024 (UTC)Reply
To add a bit to what Roy Smith described about software development: there are failures in managing it throughout the industry, particularly dealing with legacy code bases and a existing user population that generally wants all of their interactions to remain exactly the same. When the software has an associated revenue stream, there's a profit incentive to drive deadlines to be met, but when there isn't, the motivation is to get something that works implemented, as cheaply as possible. That tends to accumulate technical debt that has to be resolved later. One more developer isn't going to have much effect on these problems, which need significant resources working in concert to address. Better project management and setting of priorities is needed, to adequately plan how to transform the code base to a more sustainable state. Note there's a good possibility that would result in a decision to shed functionality currently in use that's too costly or insecure to keep in place, with a plan to re-implement some parts deemed necessary. isaacl (talk) 01:57, 24 March 2024 (UTC)Reply
That's mitigated slightly by the lack of one negative force: MediaWiki has no need to make change for change's sake, just to make Product 2024 look sufficiently different from Product 2021 that users will feel obliged to upgrade. Certes (talk) 11:48, 24 March 2024 (UTC)Reply
I'm a software engineer myself. More specifically I'm an SRE, so I'm typically responsible for the types of tasks you're talking about (server upgrades, etc). Let me give you my perspective:
For most software engineers, the work they do at their job is completely outside of their control. They do what makes their boss happy, and in turn, they do what makes their boss happy, and so on. Thankless work like regular maintenance is often dropped without the proper incentives. For some people, those incentives are the salary to work long hours, but since many American jobs in big tech pay 2x to 5x the salary WMF pays for the same role, That isn't it. Those incentives have to come from the top. An example of what that might look like is a "backlog drive" that employees are required to participate in. But that's pretty rare, because leadership is typically being evaluated on metrics like increasing revenue or visitors to the site, and technical debt doesn't push those metrics. So, asking WMF to hire more people doesn't address the problem. Those new employees, if hired, would just fall into the established system that caused the technical debt to exist in the first place. So the conversation you need to be having is: "How do we convince WMF to invest in technical debt?" I don't know the answer to that question. But focusing on hiring more people doesn't solve anything. Mokadoshi (talk) 03:31, 25 March 2024 (UTC)Reply
That's why the proposal is to hire a dev specifically to work on maintenance. Tercer (talk) 06:51, 25 March 2024 (UTC)Reply
The problem is bigger than just one person working on small bug fixes. The framework needs to be cleaned up to be more sustainable. The third-party software dependencies need to be reconciled across different extensions. This needs co-ordination across multiple development areas, and a lot of automated testing. It needs support from management to push through, rather than to just spend enough to keep the parts working. isaacl (talk) 20:49, 25 March 2024 (UTC)Reply
Been there, done that. Assigning all mainteinance work to a single engineer (or a small group of them) is a really bad idea in practice. It just obfuscates the need to have better mainteinance practices across the whole engineering organization. It isolates "mainteinance folk" from the new developments that they'll eventually move to mainteinance. It leads burns out, and then to a new "mainteinance guy" coming who starts from scratch and cannot tap into institutional knowledge about mainteinance. It is just unsustainable practice. MarioGom (talk) 10:19, 6 April 2024 (UTC)Reply
Yes, investing in technical debt is exactly what's needed, but one reason (or excuse) for not doing that is lack of people. If I gag the cynic in me shouting that any new staff would just be diverted to exciting but unnecessary new chrome, an increase in resource should make it easier to get through the required drudgery. Certes (talk) 09:14, 25 March 2024 (UTC)Reply
The key point is why hasn't the WMF already hired that one more developer, or ten, or fifty? Because it places a higher priority on spending those funds and management resources in other areas. For the development environment to truly improve, the WMF needs to change how it sets its priorities. Echoing what Mokadoshi said, that's the problem that needs to be worked on. isaacl (talk) 20:58, 25 March 2024 (UTC)Reply
Do you have a source that the Wikipedia web servers run on Debian Buster? I thought they ran on Kubernetes? Mokadoshi (talk) 19:03, 25 March 2024 (UTC)Reply
A message just came around on the cloud-announce mailing list saying that all the VPS hosts running Buster need to be upgraded in the next few months. I don't have any insight into what they're running on the production web servers, but I assume if they're migrating the VPS fleet, they're doing the same for production. RoySmith (talk) 19:14, 25 March 2024 (UTC)Reply
Thanks for that link, I'm not in that mailing list so I didn't know. I don't know how WMF runs prod so it very well may be that they are running Buster. But it's important to note that the announcement is for Cloud-VPS which is VPS hosting for the community. It would be common practice to not upgrade Cloud-VPS until the last possible minute so as to minimize disruption for the community. AWS for example does not forcibly upgrade your OS until the last possible day. Mokadoshi (talk) 19:51, 25 March 2024 (UTC)Reply
I wouldn't spend too much time and effort focusing on the Debian Buster thing. That doesn't affect end users in any way that I can see, and it is not end of life'd yet. Let's trust WMF software engineers and SREs to handle those details, and focus on things that directly affect end users. –Novem Linguae (talk) 21:05, 25 March 2024 (UTC)Reply
Where it adds to the technical debt that has to be managed is the work to figure out the third-party software stack required by the extensions used for a given Wikimedia site. I agree that it's not a level of detail that editors should be worried about figuring out, but getting the code base improved so that it's easier to work out is important for long-term sustainability of the software. isaacl (talk) 21:24, 25 March 2024 (UTC)Reply
It's in the discussion of the SVG bug I linked, where they say they will only install the bugfix when it comes with Bullseye, and link to the task for upgrading from Buster to Bullseye. Tercer (talk) 23:33, 25 March 2024 (UTC)Reply
I really don't want to speak for the WMF, but I kind of understand their logic here. One way to manage a fleet of machines is to stick with LTS releases and survive on whatever gets packaged into that. It's certainly possible to built custom installs, but once you start doing that, you're off the LTS train and have to take on a lot more responsibility and overhead. I've lived in both worlds. If you haven't, it's difficult to fully understand just how attractive sticking to the LTS can be. RoySmith (talk) 00:10, 26 March 2024 (UTC)Reply
I suspect WMF SRE would be fine with a newer version of the package, provided that there was a compelling reason (more than just small bug fixes) and a different person/team took full responsibility for what that entailed. Like if WMF multimedia team (staffed in this hypothetical future differently) basically said that this is a critical issue, we need the new version, and we are willing to take responsibility for all that entailed, it would probably happen. But that is not the situation happening here. Bawolff (talk) 15:28, 30 March 2024 (UTC)Reply
This certainly does illustrate the odd relationship that the community has with WMF. In a commercial software shop, there would be meetings between engineering and product managements. The product folks would say, "If this bug isn't fixed, it's going to cost us $X next quarter in lost revenue as customers jump ship". The engineering folks would say, "The only way we can fix this is if we go off the LTS train and that's going to cost us $Y in additional engineering costs, plus some indeterminate $Z in technical risk exposure". The two sides would argue and come to some decision, but at least both points of view would be heard and weighed.
But that relationship doesn't exist here. The customer base is the volunteer community. It's difficult to put a price tag on their labor, and even if you could, they don't have a seat at the table when it comes to making these decisions. Yeah, there's meta:Community Tech and meta:Community Wishlist Survey, but that's not quite the same thing as having the VP of sales in your face about fixing whatever bug is pissing off his biggest customer and threatening his bonus this quarter :-) RoySmith (talk) 17:48, 30 March 2024 (UTC)Reply
I think there is also a problem here in that the community lacks sufficient knowledge into how WMF does things to make effective criticism, and without effective criticism its impossible to hold WMF to account. Take this thread. The proposal is for WMF to hire a single developer to work on maintenance of MediaWiki despite the fact that there is already a lot more than one developer currently working on MediaWiki maintenance and none of the issues mentioned actually are MW maintenance issues. As a proposal it is pretty non-sensical - it is hard to even tell if these are issues the community cares a lot about or just an issue that a few people care about. The community might lack a seat at the table, which is a problem, but the community is also pretty bad at expressing what is important to it in a way you can actually do something with. Bawolff (talk) 18:17, 30 March 2024 (UTC)Reply
You're missing the forest for the trees. I could have phrased the proposal instead as "invest more resources in maintenance" without changing anything else, and your criticism wouldn't apply. Moreover, you're missing the point about "jumping off the LTS train". There is no need to update this package in isolation. If MW had been updated to Bookworm last year, or even Bullseye three years ago, they would have gotten for free the SVG bugfix. No matter which way you look at it, maintenance is lacking. Tercer (talk) 19:16, 30 March 2024 (UTC)Reply
Sure, but would it have been worth it? I also have a long list of things i would like to see fixed. The issue is everyone has a different list. I don't think there is any reason to think that hiring 10 more, or even 100 more devs would neccesarily fix the specific issues you are concerned about. Not all problems can be fixed by just hiring more people. (P.s. my criticism of, well technically this isn't a mediawiki issue, isn't neccesarily directed at you - there is no reason you should know where the boundries between different components lie. I think it is more emblematic of the meta problem where the community lacks the ability to really tell what WMF is doing and hence lacks the ability to really judge it which undermines the community's ability to be taken seriously when lobbying for changes) Bawolff (talk) 21:21, 30 March 2024 (UTC)Reply
Actually, we do have good visibility into what they're doing. The big picture of how they handle OS upgrades is on on Wikitech. And if you're willing to invest some effort looking around on phab, you can find all the details. For example, T355020 talks about upgrading the hosts that Thumbor runs on. RoySmith (talk) 22:06, 30 March 2024 (UTC)Reply
That's funny, it turns out SRE is violating their own policy by hanging on to Buster for so long. Well, I'm glad they agree with me. Tercer (talk) 22:16, 30 March 2024 (UTC)Reply
Yes, it would. Updating Debian is a no-brainer. Very likely it would take care of a large fraction of your list automatically. And it's something you have to do anyway, so there's no benefit from running archaeological versions. Keep in mind that we're talking about Debian stable here, so even the newest version is already old and battle-tested. Tercer (talk) 22:09, 30 March 2024 (UTC)Reply
I disagree. Bawolff (talk) 21:19, 1 April 2024 (UTC)Reply
  • From what I am aware of, there is a decent number of employed devs as well as volunteers, so if anything, this is a coordination problem (economical and social) more than anything else. I do have to agree on the point that despite the number of people, some important things don't seem to get done - some of it is primarily because dealing with legacy code is hard, and because hiring quality is hard (this is not to imply at all that current devs are bad) but more exactly that hiring the best devs to work on legacy code is particularly challenging (atleast without paying through the nose). The best way to resolve this is to use the donation war chest and work harder on technical evangelism + hiring on quality over quantity. --qedk (t c) 22:54, 25 March 2024 (UTC)Reply
    Perhaps WMF can fund the volunteer developers to do basic maintenance, just like the Rapid Fund and the Wikimedia Technology Fund (which is unfortunately permanently on hold)? Thanks. SCP-2000 14:53, 27 March 2024 (UTC)Reply
  • Hi all - I'm Mark Bergsma, VP of Site Reliability Engineering & Security at WMF. Thanks for this discussion and the points already raised. I'd like to help clarify a few things: at WMF we do indeed have a few hundred developers/engineers - spread over many teams in the Product & Technology department, which is roughly half of the organization. "Maintenance" is not exclusively done by a few dedicated staff, but is the shared responsibility of most of those teams and staff, for the components they're responsible for. They actually spend a large amount of their time on that, and for some teams it's the vast majority of their work. We do consider that a priority and explicitly make time & space for it (we call it "essential work"), and we aim to carefully balance it with strategic work (like bringing new functionality to users/contributors), as well as needed investments into our platform and infrastructure (e.g. our multi-year project to migrate all our services to modern Kubernetes platforms for easier development/testing/maintenance/developer workflows). Since last year, we've also made MediaWiki and related platform work an explicit priority (WE3), including the establishment of some much needed MediaWiki-focused teams again, and have an explicit annual goal to increase the number of staff and volunteer developers working on MediaWiki, WE3.2), and the new formation of our MediaWiki platform strategy. This will continue going forward (WE5 and WE6 of our draft next-year plan).
    Nonetheless, it's true that we have a big challenge sustaining the large and ever growing footprint of services, features and code we've developed and deployed over the now long history of our projects, at the large scale we're operating at. Compared to that footprint, and considering the very wide range of technologies involved, old and new, different code bases and needed knowledge and skill sets, a few hundred staff to sustain and develop that is not all that much. It involves difficult choices and tradeoffs everywhere - prioritizing between many tasks and projects, all of which seem important. It's something we care about, have done quite a bit of process improvement work on over the past 1.5 year, and have a lot more left to do on. We're planning several related initiatives (e.g. WE5.1, all KRs of PES*) as part of our next annual plan to further improve this. We're also going to communicate more about this sort of work, which has been less publicly visible before.
    But, for something more concrete right now: I've also looked into the situation of the specific issue with SVGs that you raised. That's indeed a problem with an old librsvg library version that is used by Thumbor, our thumbnailing service, which was extensively worked on last year to migrate it to our Kubernetes platform - also for easier/quicker maintenance like discussed here. Further work (including the Thumbor container OS image upgrade and some required Thumbor development for that) then unfortunately got delayed, as the development team then responsible for it needed to be disbanded and reorganized at the time - also to allow us to form the aforementioned MediaWiki focused teams. But, I'm now happy to report that the plan is for the work on the Thumbor upgrade to Debian Bullseye to start soon, in the next few weeks, which when finished should finally address this frustrating issue as well. (And yes, we do normally upgrade before OS releases go out-of-support :)
HTH! -- Mark Bergsma (WMF) (talk) 12:14, 27 March 2024 (UTC)Reply
Thanks for taking the time for such a detailed answer (it seems that Andrew Davidson had a much better grasp of the situation, and I stand corrected). I'm glad to hear that you are aware of the issues, care about them, and there are plans for improvement. In particular, I'm glad that there is a MediaWiki team again and that WE5.1 is an explicit goal. If the state of MediaWiki in fact improves (at least in the issues I'm aware of) I'll resume donating.
It's clear that the answer to my specific proposal is (a quite diplomatic) "no", but I don't mind. I care about the results, and I'm not going to pretend to know better than you how to achieve them. Tercer (talk) 18:44, 27 March 2024 (UTC)Reply
Regarding the broken Graph extension, I have written an overview in the new Signpost that hopefully sheds some light on the situation for those who haven't followed it closely over the past year: Wikipedia:Wikipedia Signpost/2024-03-29/Technology report.
Speaking personally and generally (i.e. not necessarily about the Graph situation in particular), I think people should keep their minds open to the possibility that several things can be true at once: 1) WMF does a lot more maintenance (and other technical) work than some community members give it credit for, 2) many technical issues are trickier to solve that it might appear without detailed knowledge about the situation, but also 3) WMF often has serious problems with assigning resources proportional to impact and 4) there can also sometimes be situations where engineers overcomplicate a problem and let the perfect become the enemy of the good, or lack the expertise to find the most efficient solution. Regards, HaeB (talk) 02:49, 30 March 2024 (UTC)Reply
Thank you @HaeB for your even-handed review of the Graph extension situation, and articulating some real challenges that we all face, contributors and WMF staff. Something that I remind myself daily is that most people are trying very hard to get decisions they think will have a lasting impact right, and it's true that sometimes that can lead to letting 'perfect become the enemy of good.'
I wanted to share a little of my perspective on the maintenance challenge: I've joked that this is my third 20+ year old codebase. Over those years, security, uptime and performance expectations have changed dramatically along with the Internet. We all expect software to be more safe, available and faster than ever before. Wikipedias have the privilege of serving a large user base of readers and editors, who are diverse in their geography, expectations and needs (to name just a few ways!). And often, those who work on the software are trying to keep as much functionality as possible as we modify things, so that as little as possible breaks even as the world changes around us. An imperfect analogy to what's happened might be retrofitting a series of buildings and changing their core purpose from some industry to housing. It's possible to change the purpose of a building without changing any of the internal structure, but over time, that becomes increasingly hard to live with. And now you have tenants, so you can't just send everyone away for a year to upgrade it all!
We have multiple significant maintenance projects underway. One place to hear about a part of this work is in the monthly MediaWiki Insights reports, which describe MediaWiki project progress and plans. Each language wiki and sister project wiki is a separate deployment that also supports a distinct set of extensions, some of which were created and deployed many years ago, under pretty different circumstances than we face today. Following the insights reports may help those interested learn more about what it takes to untangle years of uncoordinated decision making for our most critical software, and what it will require for us to establish a platform that is supportable on a free and open internet by dedicated staff and volunteers, and still enable an open and distributed model for all contributions.
Having volunteers who understand our challenges and then help us think through how to get things right together is super helpful. Thanks again and please keep sharing your thoughts. SDeckelmann-WMF (talk) 19:42, 1 April 2024 (UTC)Reply

Preannouncement of upcoming Movement Charter conversations edit

I am Kaarel, support staff of the Movement Charter Drafting Committee, working with the Wikimedia Foundation.

I am writing here to let you know in advance that the full draft of the Movement Charter will be published on April 2nd, 2024. This will kick off the community engagement period from April 2nd to April 22nd. Perspectives from the English Wikipedia community would be very valuable for the conversations.

For context, the Movement Charter is a proposed document to define roles and responsibilities for all the members and entities of the Wikimedia Movement, including to lay out a new Global Council for movement governance.

Everyone in the Wikimedia Movement is invited to share feedback on the full version of the Charter draft – this is the last chance to propose improvements before the Charter draft is updated for the ratification vote in June 2024.

Since the last feedback round the drafts have gone through notable changes. I hope many of you will still find it worthwhile to review the drafts and share your perspectives to inform the final version of the text that will be ratified.

How to share your feedback?

Read the Committee's latest updates for more information. I am truly grateful for your kind attention!

On behalf of the Movement Charter Drafting Committee, --KVaidla (WMF) (talk) 07:38, 26 March 2024 (UTC)Reply

Please remember, when attempting to tie-in and enforce any specific rulings or wording of the document, that this project is Wikipedia (specifically English Wikipedia) and not Wikimedia or an entity called "Wikimedia movement", and its editors are called Wikipedians and not Wikimedians (although some Wikipedians may also want to self-identify as Wikimedians). Thanks. Randy Kryn (talk) 12:27, 27 March 2024 (UTC)Reply
Thank you for this feedback! I would like to understand better the point you are making. The Wikimedia Movement Charter is a high level document defining the future roles and responsibilities of those comprising it. As defined here, on English Wikipedia, "The Wikimedia movement is the global community of contributors to the Wikimedia projects, including Wikipedia." The article further elaborates that "the Wikimedia community includes a number of communities devoted to single wikis", followed by a list, including Wikipedia communities.
In my perspective, this means that following the definitions also in respective article on English Wikipedia, as curated by English Wikipedia community, Wikipedians are part of particular Wikipedia community, who in turn are part of the global Wikimedia community. It does not seem reasonable in a Charter-like high level document to mention all the projects separately, so the term Wikimedia is used. At the same time, I do hear your point about the choice of the term might feel "alienating" (my interpretation, please correct or specify, if not correct) for the contributors of particular projects. What is your positive proposal for terminology considering this?
Thank you for your very kind attention given to the matter! --KVaidla (WMF) (talk) 15:46, 28 March 2024 (UTC)Reply
Hello KVaidla (WMF). The point is quite simple. Wikipedians are people who have edited Wikipedia, the flagship project that WMF was organized to fund and protect. Wikipedians can individually choose to identify as Wikipedians, Wikimedians, or both (irregardless of the language you quote above, remember that Wikipedia cannot be used as a source) and are different groups of people when they identify or identifying them. WMF's funding is primarily based on donations that people think they are donating to Wikipedia when, in fact, Wikipedians have very little to do with deciding where and how much such funding is allotted. Wikipedians, a dictionary term, even used to have an annual award named for them. Wikimedian of the Yearl was named 'Wikipedian of the Year' from 2011 to 2016, when an obvious opportunity not taken arose to create the second award while keeping the first intact.
Since Wikipedians are a stand-alone grouping, they may choose to belong or not belong to what you call the 'Wikimedia movement'. But any rulings or direction which dictates WMF (or "Wikimedia movement") policy onto them, without regard to the separation of the two, should not stand as doctrine or overriding policy but simply as suggestion. Funding of Wikipedia projects, conventions (our Wikimedia conventions, both worldwide and local such as the North American convention, should rival other major corporate and professional conventions in terms of facilities, banquets, sponsorship, etc., with a much greater extension of WMF funding, upwards of 500 scholarships to each, etc. as a major opportunity to celebrate the volunteers), and new ideas should be taken for granted in WMF's yearly budget, with WMF asking "Thank you, what more can we do?". In any case, the elephant in the room is that Wikipedia is the popularly known entity which "brings in" the donations for the movement and WMF operations, and instead of erasing the term 'Wikipedian' (as was done with the annual award) that separation, as well as the partnership and much fuller funding, should be further recognized and encouraged. Thanks for your reply above, and for your work supporting the projects. Randy Kryn (talk) 15:37, 30 March 2024 (UTC)Reply
I am proud to be a Wikipedia editor. I don't really know or care about this "Wikimedia movement" which some marketing genius seems to have plucked out of thin air recently. I associate the term "movement" with promoting some cause or other, whereas I try hard to do the opposite by editing neutrally. If membership of this "Wikimedia movement" is optional, then I opt out. If it's a mandatory condition of remaining a Wikipedia editor, just let me know, and I'll move on and leave you to write your own encyclopedia. Certes (talk) 22:31, 1 April 2024 (UTC)Reply
The “membership” like it pleases you to frame it is optional but certain policies are not. For camper, you may not personally attack other users or perform undisclosed paid editing. Ymblanter (talk) 11:40, 2 April 2024 (UTC)Reply
I haven't done either of those and don't intend to, but I consider myself answerable to the community and not the WMF. I'll continue to behave in a way I feel is reasonable, the WMF (unfortunately) can decide which of us it allows to edit, and I trust the community to react appropriately to any poor or unexplained decisions as we have occasionally had to do in the past. Certes (talk) 11:59, 2 April 2024 (UTC)Reply
Thank you, Randy Kryn, for taking the time to elaborate your point more. clearly. I believe I understand the perspective better and it has sparked further conversation, which can be helpful.
As to separation between Wikipedia communities and Wikimedia movement, I believe that project autonomy and agency of project contributors in policy-making continue to be important aspects of how we function. At the same time, there are already global policies, like the Terms of Use in place, that basically legally enable the functionality of all the projects. The Wikimedia Movement Charter drafting work is based on a hypothesis that there are matters that can be defined universally, so we do not need to revisit these matters project by project. Also it is about having the rights and responsibilities of different stakeholders set in writing for clarity. It would be interesting to hear your further thoughts and insights in relation to what has been suggested in the Movement Charter draft (which is now been published). Thank you again for your time and sharing of your perspective! --KVaidla (WMF) (talk) 12:58, 4 April 2024 (UTC)Reply
Thanks KVaidla (WMF). I think my statement above presents my personal concerns, mainly about full funding of Wikipedia/Wikimedia volunteer projects and the worldwide and regional conferences (where 257 scholarships are given, for example, expanding that to 700, or 800, and assuring that the conferences rival the best corporate or non-profit conferences in keeping with Wikipedia's reputation and, importantly, to thank some of the volunteers who provide both the work product and the incentive for donors) could easily become a key priority of WMF funding. I've been saying that the world's billionaires will, at some point donate up to a billion dollars to Wikipedia and Wikimedia associated projects (Wikipedia's perceived yearly worth per the EMO - the Elon Musk Offer). For both then and now it would be nice to assure that the volunteers fully participate, including widescale funding-decision participation, in everything that dedicated knowledge-sharers can think up and accomplish. Randy Kryn (talk) 14:46, 5 April 2024 (UTC)Reply

The movement edit

The phrase "Wikimedia movement" is not a recent invention, though it has received more attention in the past couple years. We individual Wikians of course are free to call ourselves proudly as we will but I think the failure a few years ago to rename the outlying and ancillary sites to "Wikipedia [something]" was unfortunate as it would have helped in public recognition. It would help in recruiting; almost all my students have been attracted to our teaching sessions in hopes of writing a new article when really, it would usually be easier for them to learn how to add a new picture. For my own editing, for a decade or more I have been putting about as much effort into Wikimedia Commons and Wikidata as into English Wikipedia. Those sites do a few things but mainly they supply pictures and data (geographical coordinates are one of my specialties) to the hundreds of Wikipedias including English. Even without a common name, I figure there ought to be a common policy structure and general set of guidelines for us all, so the WMF staff don't have to decide so many things for us. Naturally there will be an interest in adding matters that are mainly the concern of only the encyclopedias or of only few of the various other autonomous member sites, who each ought to handle such things by their already established methods. There will be disagreements on such questions, but I'm confident our usual methods of drawing a consensus will prevail. Jim.henderson (talk) 00:16, 4 April 2024 (UTC)Reply

  • In short: I'm curious whether you're saying you believe that a difference of one letter (in some cases) constitutes a real detriment to the development of Wikipedia's sister sites? "Wiki" is already very widely associated with Wikipedia by the public, does "Wikipedia Data" being a thing people hear about and are surprised it exists seem much more unlikely than with "Wikidata"?
  • I really disagree that even in effect Commons's main function is to act as Wikipedia's image repository. Its use is ubiquitous.

I figure there ought to be a common policy structure and general set of guidelines for us all, so the WMF staff don't have to decide so many things for us.

  • I don't really think this structure makes sense. WMF doesn't really make that many choices for us—or at least, I feel part of the process for just about every choice I would like input on. I'm not sure what exactly you would like centralized, the current model seems to work. Do you have specific areas of policy or administration in mind?
Remsense 00:39, 4 April 2024 (UTC)Reply
I think it's important to remember that there isn't a "we", strictly speaking. Not all of us are here to be part of a movement or whatever, or even editing for the same reasons; some just want to fxi a tyop that bugged them, others to add information about their favourite volcano. Jo-Jo Eumerus (talk) 07:18, 4 April 2024 (UTC)Reply
I have a great respect for sister projects. I refer to Wiktionary regularly, and I use Commons unconsciously most times I read a Wikipedia article with images, though I've contributed very little to either. I use Wikidata occasionally, and was briefly an enthusiastic contributor; I would use it much more if it had an intuitive user interface. None of this makes me part of a "Wikimedia movement", which I see largely as a political stance shared by some but not all editors. Certes (talk) 09:08, 4 April 2024 (UTC)Reply
The fact that you think the other projects are nothing more than ancillary sites that should be rebranded as under the Wikipedia umbrella is a very, very bad sign. Cremastra (talk) 22:04, 5 April 2024 (UTC)Reply
  • I completely understand that Wikipedia editors have a very Wikipedia-centric view of the broader Wikimedia community. It would be nice if everyone had the interest to learn more about the other projects, and how they are used, but I understand that it's not a priority for most Wikipedians, regardless of which language Wikipedia they choose to edit. But I'll just throw in a few tidbits:
  • For many mainstream media sources that use images, Wikimedia Commons is often a go-to resource. There isn't a day that I don't find a Wikimedia Commons credit on the series of mainstream media sources I look at.
  • Wikidata is used widely as a data aggregator by many other not-for-profit resources.
  • Some language editions of Wikisource are the largest repositories of open-source historic documents and literature in those languages
  • Some language editions of Wiktionary are having a major impact in preserving dying languages

The 339 Wikipedias, with their collective 62+ million articles, are indeed the major drivers of the Wikimedia family of projects; the largest Wikipedias, especially the English one, have deservedly gained a reputation for accuracy and neutrality in providing the entire world with information. That doesn't mean that the "sister projects" don't make a valuable contribution to the sum of all human knowledge. I will never fault anyone for choosing to focus on one specific Wikimedia project, or even one small aspect of a specific project. We're all better for those contributions. Risker (talk) 08:39, 4 April 2024 (UTC)Reply

  • Fearful of being caught beating a dead horse, I'll speak once more and hope to be disciplined enough to hold my peace. Yes, one little letter can indeed make a difference. Besides the local Wikimedia Chapter I am a member of local clubs concentrating on bicycling and astronomy. I'm always nattering to them about how we make Wikipedia and sometimes they actually listen. A couple times, though they did not surrender to my efforts to get them to edit an article or two, they decided that a money contribution would be a good thing. "Eh? The form said 'Wikimedia Foundation'. Is that the same thing, or did someone else sneak in with some kind of Internet scam?" These people had me to make the explanation, unlike most outsiders.
  • A common brand can't help us thousands of insiders to understand what we're doing; like other big busy communities we are too complex to be understood with just a few simple labels. However, that's not what brands are for. Brand names are to promote instant understanding among the mass of outsiders, in this case their understanding that this is a big complex of websites with a common theme. We have different detailed policies to handle our specialties, but we are all about Wikipedia's original aim of promoting universal knowledge by organizing it for accessibility.
  • I have uploaded thousands of pictures to Wikimedia Commons, and I know of a dozen that are used in news publications, books, and other works. Probably hundreds more are so used; we don't require to be given notice. The usual credit, when given, is "From Wikipedia" or "From Wikimedia Commons" or "by Jim.henderson via Wikimedia Commons". I figure, out of the small minority of their readers who actually wonder where the pictures came from, "From Wikipedia" is the only one that isn't mysterious. Yes, knowledge professionals, most often, know what they are doing but they hope their product will be read by many more people than are composing it; same as we do. They don't expect their readers to learn what it takes to be a knowledge professional.
  • Wikidata is indeed used by many knowledge professionals. I am most familiar with the work of librarians since I work with them often, and have a lunch appointment tomorrow with a relative who's in that line of work. They are familiar with the different workings of OCLC, Worldcat, LoC and various local or specialized catalogs, and many of them also use Wikidata to help find their way through and among the others. Haha, a couple years ago I looked at the Wikidata item about me, and it showed my LoC Authority Control Number. What, has my good work come to such high prominence? No, that was another Jim Henderson, so I corrected it. Wikidata is full of dirty data like that, some of it imported from massive outside databases a century old or more. Not a big problem since WD serves people in the database business who are aware. I have also worked with art museum archivists and have no idea whether their old catalogs have as many errors as the databases for community gardens in Brooklyn or defunct restaurants in The Bronx. Hmm, perhaps I have wandered but anyway yes, the question of what brand name Wikidata ought to use is a lot less important than for Wikiprojects that serve a wider direct audience.
  • Oh-oh, it seems this paragraph on WD has revealed that I've already gone overboard. So, I'll drop my stick and carefully back away from the dead horse. Jim.henderson (talk) 01:13, 5 April 2024 (UTC)Reply

The full Movement Charter draft awaits your review on Meta edit

Hello again! I am following up on the pre-announcement from the previous week to let you know that the full draft of the Movement Charter has been published on Meta for your review.

Why should you care?

The Movement Charter is important as it will be an essential document for the implementation of the Wikimedia 2030 strategy recommendations. Participating in the Charter discussions means that you ensure that your voice is heard and your interests are represented in shaping the future of the Wikimedia Movement. As the English Wikipedia community is the largest of the Wikimedia movement, it is essential to have the perspectives from your community presented in the global conversations. I hope many of you will find time to provide feedback, share your thoughts and perspectives!

Community Engagement – April 2nd to April 30th

The Movement Charter Drafting Committee (MCDC) cordially invites everyone in the Wikimedia movement to share feedback on the full draft of the Movement Charter.

Let your voice be heard by sharing your feedback in any language on the Movement Charter Talk page, attend the community session today, on April 4th at 15.00-17.00 UTC, or email I will also be monitoring conversations on this talk page, to bring back the summaries to the ongoing global conversations.

You can learn more about the Movement Charter, Global Council, and Hubs by watching the videos that the Movement Charter Drafting Committee has prepared. Read the Committee's latest updates for more information about the most recent activities from the Drafting Committee.

Thank you again for your time and kind attention! I look forward to your input and feedback. Have a wonderful month of April! --KVaidla (WMF) (talk) 13:07, 4 April 2024 (UTC)Reply

Unified enwiki response to the charter edit

In votes like these a significant issue is that interested editors do not have the time or wherewithal to properly assess the issues or candidates presented and so abstain from the vote. I propose that we attempt to address this, by having more engaged editors consider the proposal carefully and, in consultation with the community though an RfC, issue a recommendation either to support or oppose the change. Specifically, I propose a three-stage process:

  1. A pre-RfC discussion where we will write a neutral summary of the proposal.
  2. An RfC where we will:
    1. !Vote to approve the summary and its dissemination
    2. !Vote whether we should encourage eligable enwiki editors to vote for or against the change
  3. Assuming the summary is approved, a mass message to all eligable enwiki editors providing it. Further, assuming there is a consensus either for or against the change, a recommendation to the editors that they vote in line with that consensus.

Stage one should probably begin soon, in time for a RfC in May; first, however, I wanted a brief discussion of the general idea. BilledMammal (talk) 02:41, 7 April 2024 (UTC)Reply

Thanks for your comment, BilledMammal. This isn't a bad idea, but it is worth noting that the draft charter will be revised in early May following this current feedback round. Although the MCDC (of which I am a a member) does not anticipate making really large changes, I think it would be reasonable to assume that the final version is going to have at least some differences from the current draft. Would it make sense to create a feedback page on this project as a place where interested enwiki editors could flesh out their opinions before the final revision is made? I'd hate to see people investing a lot of time reviewing a draft and proposing a project-wide opinion in an RFC-type format, based on a document that we know will change. There is something to be said for having a local page for comments and suggestions for improvement (and please yes, if someone thinks X is a bad idea, propose an alternative) as long as there's a link to it on the Meta page so that the MCDC will be well-informed of the discussion on this project. (For that matter, it may be a good idea for other projects, and I'm pretty sure some of them are thinking about this too.) Risker (talk) 03:17, 7 April 2024 (UTC)Reply
That's a good point, but we also need to consider that it will take time for the RfC to run. I think we should start drafting the summary based on the current document, and then make any updates that are necessary to align it with the May changes and start the RfC a few days after it is released.
I would also agree that creating a local page where editors can make comments and suggestions for improvements would be useful, although I would suggest just using this page as it isn't as busy as the other village pumps and thus an extensive discussion of the proposed charter won't disrupt other discussions. BilledMammal (talk) 04:59, 7 April 2024 (UTC)Reply
I think mass messaging every eligible voter WP:ACE style might be too many people. Perhaps a watchlist notice, or pinging rfc participants, would be a good compromise. –Novem Linguae (talk) 16:30, 7 April 2024 (UTC)Reply
I don't think that the vote to ratify this charter is less important than the vote to elect the Arbitration Committee. BilledMammal (talk) 16:32, 7 April 2024 (UTC)Reply
Thank you for this initiative, BilledMammal, to approach the Movement Charter conversations in a constructive way! For reference, the timeline for the steps can be found here on meta and you are right, the time is of essence. It has been already pointed out on the meta discussion page that the review of the Charter would benefit from additional contextual materials for informed decision-making. As a supporting staff member to the MCDC I will see what I can do, yet it might take some time. If there are priority areas for further context in the English Wikipedia community, please let me know, so I can focus my work around that and hopefully have respective content available earlier. Also let us know, if we can support the discussions around the Charter in other ways. Looking forward to hearing the perspectives and seeing good participation from en.wp community! --KVaidla (WMF) (talk) 12:47, 9 April 2024 (UTC)Reply