Proposals for closing projects/Move Beta Wikiversity to Incubator 2

From Meta, a Wikimedia project coordination wiki
Jump to: navigation, search
This is a new, cleaner version of the previous proposal (closed on procedural grounds), per comments by SPQRobin, Lsanabria, and others.

This is a proposal for closing and/or deleting a wiki hosted by the Wikimedia Foundation. It is subject to the current closing projects policy.


The proposal is currently open for discussion by the community.


Proposal: Move Beta Wikiversity to Incubator, incorporating good ideas + discussion spaces from beta.wv into Incubator. Specifically:

  • Create a landing page for Wikiversity on Incubator: with non-incubation discussions from beta.wv as subpages. (Each sister project would benefit from this.)
  • Migrate the ~500 local images from beta.wv to Commons. There are no fair use claims, no recent media uploads.
  • Improve RC and search on incubator so people can view/search only WV projects. Similar to the current restriction by project, but for wv:all. (Each sister project would benefit from this.)
  • Migrate each language Wikiversity from beta.wv to Incubator.

Process: I pinged the most active past participants and b.wv editors. Feedback from current bureaucrats Samuele2002 and Crochet.david should be especially attended to: if they are sure the Incubator can not support what they need, we & the incubator maintainers should understand why & what that means for other sister projects. 18:04, 3 July 2017 (UTC)

Summary[edit]

Past proposal
Move Beta Wikiversity to Incubator 1 (2013 discussion)

The 2013 proposal changed focus while discussion was underway, and was closed after a meandering history. SPQRobin wrote in closing "the proposal needs to be clearly defined or have a clearly defined decision-making process." Lsanabria wrote "open a new one to discuss the idea, providing a summary of topics discussed".

This proposal is for a clean merge of the two wikis, preserving current tools and wiki functionality. In the process of discussing it, we will hopefully sort out confusions about the differences b/t the wikis, identify what Incubator can learn from beta.wv, and decide whether or not Incubator is a good place to incubate new Wikiversities.

If there is support for a well-defined merge, we can then define details of what that could look like, including improvements to Incubator, and then have a fixed-duration discussion (a month?) for people to review and approve the final proposal.

Arguments for the proposal[edit]

  • This is why Incubator exists.
  • Translations and activity-tracking are clean and easier there; approval should in theory be faster.
  • Separation and duplication is wasted work: spam, SWMT, tech changes would be slightly easier for a single incubator site. [but see arguments against re: utility of separation]
  • The main: and file: namespaces on beta.WV are messy (everything in once place), hard to migrate from, and not well maintained. (example)
  • A developing wiki benefits from support from other sister projects, other groups working in the same language, and the tech+translator network. All are engaged w/incubator. beta.WV divides out only those interested in Wikiversity.
  • Discussions about global topics, and general activity, benefit from a critical mass of active users.
  • Custom features [like sister-project-only search, RC, coordination pages] should be added to incubator, as an option for all sister projects. In the medium term, this is easier to maintain than separate beta-wikis for each project.
  • Provides LanguageConverter support for e.g. Chinese and Serbian, which is actually supported on Incubator (try Wy/sr in both sr-ec and sr-el)

Arguments against the proposal[edit]

  • Migration is work. Who wants to waste time on it? Current setup: notbroken.
  • The active communities should decide. (Starting with beta.WV, including localizers, small-language groups, SWMT, and devs of incubation-tools)
    beta.wv has activity -- Recently hi.wv has activity, and is being approved. Jayprakash worked on that. zh.wv and he.wv have also been active. (See this curious page of number theory.) A couple short policy discussions happened on their Babel forum Crochet.david has been handling high-level requests and could weigh in here.
  • Separation and duplication can be good.
    WV-only RecentChanges. [incubator supports this for a single language-project. Doing that for a whole set of sister projects (e.g., wikt/*), isn't currently working.]
    WV-only search results. [incubator lets you set a project-preference that affects searches for a single language-project. No way yet? to search across wikt/* or */de.]
    NB: global announcements go to beta.WV's Babel (which everyone uses to chat), but go to a 'WM News' page on Incubator w/ no sidebar link. Is community space on beta.WV more effective?
    These announcements are made to Incubator:Wikimedia news. However, that's not obvious, so I do need to put a notice to that effect on Incubator:Community Portal. Thank you for bringing it to my attention. StevenJ81 (talk) 13:53, 6 July 2017 (UTC)

Arguments refuted, or neither for nor against[edit]

  • All the Wikipedias and all the Wikiversities are free to use fair use images, including Beta Wikiversity and any potential new ones developing in BetaWV --Marshallsumter (talk) 02:37, 7 July 2017 (UTC)
    Not all Wikipedias and Wikiversities are allowed to use fair use. Only couple of Wikimedia projects are.— Danny B. 17:31, 10 July 2017 (UTC)
    This is not a current concern. Beta.WV has deleted new fair use images (since 2014) and migrated most images to Commons. No current projects are asking to upload such images.
    Neither beta.WV nor Incubator seems to have an EDP. (Not too surprising, as young projects rarely have much media.) If any new project wants fair use images, Incubator could allow it. As soon as a project gets its own wiki they can implement their own EDP. SJ talk  02:47, 11 July 2017 (UTC)
  • The ability to search all WMF projects is coming to Wikiversity. While this argument may not seem for or against the proposal, the freedom to research other WMF projects is especially beneficial to any Wikiversity and its contributors. Putting BetaWV in an incubator could and probably would stymie research. --Marshallsumter (talk) 14:44, 7 July 2017 (UTC)
    The discussion at user talk page can dispel misunderstandings and put some matter to rest. --George Ho (talk) 06:09, 8 July 2017 (UTC)
    BetaWV is already an incubator. The question is whether to share an incubator or keep it split up.
    Cross-wiki search is coming to all projects. The question is how to implement it will likely be similar on Incubator and on beta.WV. SJ talk  02:47, 11 July 2017 (UTC)
  • Original research can only be conducted and performed on Wikiversities not on incubators --Marshallsumter (talk) 02:45, 7 July 2017 (UTC)
    Incubator has no policy against OR, and Wikinews projects (defined by original research) have had no trouble on that account.
  • We could modify our Incubator policy to allow original research if you re-consider to support us. --Liuxinyu970226 (talk) 09:38, 7 July 2017 (UTC)
    • I appreciate your interest in original research! Unfortunately, others who "run" incubators are more likely inclined toward a Wikipedia view of synthesis or OR than allowing it to be encouraged. --Marshallsumter (talk) 14:44, 7 July 2017 (UTC)
      • Besides "citation needed" for your previous statement, there is absolutely no issue to set the Incubator policy regarding the original research to something like "Original research is not permitted with the exception on pages with Wv/ prefix" to match the fact that Wikiversity allow original research.
        Danny B. 17:28, 10 July 2017 (UTC)
        • What about WikiJournal, Danny B.? If it becomes a full-fledged project, how will the Incubator handle WikiJournal's original content? --George Ho (talk) 17:50, 10 July 2017 (UTC)
          Incubator has no opinion on such things: it exists to host projects that are just getting started, whatever their policies. To my knowledge Incubator has never yet incubated a new kind of sister project [like WikiJournal] -- but I don't see why that couldn't happen. (Meta has also hosted such demo-sites in the past.) SJ talk  04:02, 11 July 2017 (UTC)
  • I think it is advantageous to keep beta.Wikiversity for a variety of reasons: to help create new Wikiversity and to have a special project could stimulate the creation of new Wikiversity (also note that Wikiversity is The project with less language versions). Plus having beta.Wikiversity is easier to monitor the creation of new pages in the various wikiversity hosts on beta.Wikiversity also the beta function is to be a reference point for the communities of various Wikiversity in any language (eg a few months ago there We are coordinated and decided there to edit the Wikiversity logo) both for proposals that relate to the project and to help individual wikiversities in introducing into original searches. In addition, beta.Wikiversity would be the place to test experimentation on the project or to try extensions (eg Education_Program, Quiz Mooc...) all this if beta.Wikiversity was moved to an incubator would not be possible. --Samuele2002 (Talk!) 22:07, 9 September 2017 (UTC)
  • (Sorry if I could repeat arguments that have already been said but it is to divide by points)Beta.wkkiversity performs an incubating function of the new Wikiversity by paying attention to the needs and needs of the individual --Samuele2002 (Talk!) 22:15, 10 September 2017 (UTC)
  • Beta.wikiversity allows new wvs to learn and test the Wikiversity's own functionality (Education_Program Mooc quiz extensions...). --Samuele2002 (Talk!) 22:15, 10 September 2017 (UTC)
  • beta.wikiversity is the place where it is possible for the new wv to develop original researches and among the objectives of beta.wikiversity there is also the one to encourage them. In addition, beta.wikiversity promotes the possibility of allowing the development of RO in the various wikiversity. --Samuele2002 (Talk!) 22:15, 10 September 2017 (UTC)
  • beta.wikiversity assumes the role of coordination point of the various wikiversity on decisions concerning the project (as for the logo). --Samuele2002 (Talk!) 22:15, 10 September 2017 (UTC)
  • In addition wikiversity is the project with fewer language versions and maintaining beta.wikiversity encourages the creation of new Wikiversity. --Samuele2002 (Talk!) 22:14, 10 September 2017 (UTC)

Interested recent participants[edit]

@Crochet.david:, @Jayprakash12345:, @Samuele2002:, @Green Zero:, @George Ho:, @Lsanabria:, @Millosh:, @SPQRobin:, @Sj:, @Liuxinyu970226:

Please ping other active editors on beta.wv, and people interested in cross-project collaboration & features on the incubator.

I am here requesting some user. Who maked Hindi Wikiversity agenda for their purpose. The of langcom commented are " Langcom has reviewed the concerns and found them unfounded. So please continue with the creation of the wiki." So please don't make Hindi wikiversity as agenda. All the things have negative and positive views. But only described negative point is not right. Thanks for being here.-Jayprakash12345 (talk) 06:52, 25 July 2017 (UTC)

Comments[edit]

Short comments only, please. To respond or elaborate in detail, please use other sections below.
  • Support I support this proposal On the condition that quiz ext should be installed and Bot should be retain. bur and sysop flag can convert in 6 month sysop flag For active user. Thanks-Jayprakash12345 (talk) 13:27, 3 July 2017 (UTC)
  • Comment Based on discussion so far: I think doing this properly will help speed up new WV development, make maintenance + language support easier, and improve Incubator. It's important to carry over useful features of beta.wv (also useful for other non-wv projects!) And as Jayprakash says, user flags should be converted. But I want to better understand the issues Samuele and David raised in the past, to see if the Incubator could include a community space that works for them. SJ talk  16:02, 3 July 2017 (UTC)
  • Support There is no reason why Wikiversity should have some separate Incubator. For general collaboration about the project, that's why Meta exists. In addition to some bad pages and low level of activity, they have actually tried to graduate the Hindi Wikiversity which contains about 500 pages, 100 of which are just Linux code with no commentary. This is after five and half years. A project like this would never be released from Incubator, nor should it be. Please close Beta, fold appropriate content into Incubator and Meta, and lock the domain--it is serving no purpose and it is graduating projects which are not appropriate. Up until now, you could claim that there was some benign neglect happening where it doesn't cause any harm to have it, so why bother but I think there is demonstrable and tangible harm in promoting projects which are unsustainable. It is actively working against the purposes that it is supposed to serve. Has the community been properly notified? I don't want this knocked out on a technicality (again). —Justin (koavf)TCM 18:11, 3 July 2017 (UTC)
    • Yes, notice was placed. One nice thing about Incubator is, once you've set your project preference, you can focus on that wiki. So there's no need to become an independent wiki before there is an active community. SJ talk  16:10, 4 July 2017 (UTC)
  • Comment - Feel free to comment here, Marshall. Thanks. --George Ho (talk) 23:13, 7 July 2017 (UTC)
  • Oppose Before these proposals to close BetaWV and distribute it to Meta and Incubator, some 14 Wikiversities became open. Since these closure proposals began only 2 became open or have approval ongoing! There are 21 more Wikiversities eligible for approval. The biggest effect of these repeated closure proposals has been to discourage additional eligible Wikiversities from reaching for approval before opening. Apparently, it hasn't occurred to the closure proposers that like the additional language Wikipedias, Wiktionaries, and the many others, most participants also contribute money to the WMF for the pleasure of enjoying each project in their beloved languages and translating among them. --Marshallsumter (talk) 00:50, 8 July 2017 (UTC)
    • Or, one could say that it was because of Beta Wikiversity itself that no projects were promoted. --Rschen7754 00:58, 8 July 2017 (UTC)
      • I took a look at when the two proposals occurred. Closure of Beta Wikiversity began 27 May 2008, during that month three Wikiversities opened, that proposal was closed on 4 December 2013 before the 15th Wikiversity was open in February. So the first closure proposal may have been discouraging but did not stop the openings. The second one Move Beta Wikiversity to Incubator 1 began 25 June 2013 after the Korean Wikiversity opened and lasted to 3 July 2017, but the Hindi Wikiversity was approved in June 2017. Both periods disprove "it was because of Beta Wikiversity itself that no projects were promoted." With 21 more eligible for approval yet only one approved, the 22nd, it should be clear that this repeated effort here to close Beta Wikiversity is the sole negative force of consequence. --Marshallsumter (talk) 20:06, 8 July 2017 (UTC)
        • Hi Marshall, longer comments might fit better in the questions section below. I've responded there. SJ talk  02:20, 11 July 2017 (UTC)
  • Support There's no reason to have a separate site for just Wikiversity, and in fact there are legitimate concerns that this sets up an unhealthy isolation from the rest of Wikimedia. --Rschen7754 00:58, 8 July 2017 (UTC)
  • Neutral - I'm close to "opposing" the proposal. The list of incubating wikis is humongous, and I'm not convinced that the Incubator is well equipped to graduate a Wikiversity site. However, I'm convinced that creating another Wikiversity would take years. Time span between creating Korean and Hindi Wikiversity was four years. Jayprakash12345, Sj, and Rschen7754, what will we do with the load of incubating wikis, especially when Beta Wikiversity will be merged into the Incubator? Also, Jay, how will the Incubator community accomplish your suggestions?

    Moreover, since 2014, no more than five wikis have been created annually, i.e. graduated from the Incubator by the language committee and the Phabricator developers. I'll soon notify some more projects for their opinions on this re-proposal. --George Ho (talk) 21:37, 8 July 2017 (UTC); amended, 21:58, 8 July 2017 (UTC)

    • "Moreover, since 2014, no more than five wikis have been created," ... what? I'm counting 12. --Rschen7754 21:39, 8 July 2017 (UTC)
      • Oops. I amended my comment, Rschen7754. Thanks for pointing out my err. --George Ho (talk) 21:58, 8 July 2017 (UTC)
        • Switched to weak Oppose after finding out that the list of incubating Wikiversities is also huge. I don't see a real benefit of overwhelming Incubator community with a load of incubating Wikiversities. --George Ho (talk) 02:03, 9 July 2017 (UTC); Switched to "oppose". --George Ho (talk) 04:19, 9 July 2017 (UTC)
        • Moreover, the last Wiktionary created was 2013 and Venetian. The first Wikinews graduated from the Incubator as of this year and is Dutch. The last Wikiquote created was 2013 and Sanskrit. Wikivoyage was last created last year and is Finnish. Outside the Incubator, Wikisource was last created in 2013 and Assamese. Seems that most of graduated wikis are Wikipedias in non-English languages, especially this year. I wonder whether those at Incubator can spend energy on creating more sister projects. --George Ho (talk) 04:18, 9 July 2017 (UTC)
        • Also, as of now, not one Wikibooks site graduated from the Incubator. --George Ho (talk) 04:21, 9 July 2017 (UTC); struck out my misassumption. --George Ho (talk) 16:41, 10 July 2017 (UTC)
          • That is an incorrect statement. In 2010 the Limburgish Wikibooks got created from the Incubator. --OosWesThoesBes (talk) 12:49, 10 July 2017 (UTC)
            • Oh... I didn't see "wb/li" frankly, OosWesThoesBes. My bad. Still, not one Wikibooks site for seven years after Limburgish? --George Ho (talk) 16:41, 10 July 2017 (UTC)
  • <sigh>Again?</sigh> Strong support. Incubator is a site to put pages of wannabe-wikis to, Meta is the communication hub. No need for extra site which duplicates existing functions/wikis. Not even speaking about naming issues when names are similar in different languages and other problems which the beta Wikiversity has regarding starting new wikis. Incubator can actually help it to evolve better and faster.
    Danny B. 16:50, 10 July 2017 (UTC)
    • I don't know, Danny B.. I was told that the LangCom handles approvals and closures of local wikis. As a result, since 2014, several of them passed annually, most of them Wikipedias. Meanwhile, most of the incubator wikis sit there almost dusting. --George Ho (talk) 17:56, 10 July 2017 (UTC)
    Comment Comment I am quite upset that the same mistake happens here like in previous discussions on this topic - the discussion becomes messy and arguments for both views are quite hard to track instead of being clearly summarized and structured. No wonder why previous discussions took ages and went nowhere... :-/
    Danny B. 17:24, 10 July 2017 (UTC)
    {{sofixit]} ;) Seriously: I'm trying to keep the intro clean; if you can help refactor it, please do.
  • I placed my comments below, under an appropriate question, because I think there are a lot of misconceptions here. StevenJ81 (talk) 19:44, 10 July 2017 (UTC)
  • continue supporting I said reasons time and time again, I don't wanna repeat them. --Liuxinyu970226 (talk) 07:45, 11 July 2017 (UTC)
  • Support. I support this proposal because, as it was mentioned in the summary section, that is why the incubator exists, it will be a better use of shared resources and the Incubator has better namespace management to support multiple language codes. Any inter-language coordination can be done in Meta, which is also one of the reasons why Meta exists. Also, I think it will be easier to monitor projects to identify when they get the level of activity required to be promoted to independent projects. Regards, Lsanabria (talk) 15:12, 24 July 2017 (UTC)
  • Oppose see this my edit for my oppose. --Samuele2002 (Talk!) 22:11, 9 September 2017 (UTC)
  • Oppose as per Samuele2002. --pegasovagante (tell me) 14:31, 12 September 2017 (UTC)
  • I'm personally remaining neutral on this. However, it is abundantly clear to me that there is no consensus to support this proposal. So I strongly suggest that this proposal be closed as not successful, and that there be a two-year moratorium on reintroducing it. StevenJ81 (talk) 15:20, 12 September 2017 (UTC)

Questions[edit]

Can we update incubator to include features + tools beta.wv currently has?[edit]

For most technical aspects, it looks like the answer is yes: options could be the same. The primary difference is a social one: having a separate community provides a different sense of what is going on; why people come to the site; what local norms and standards are for welcoming and supporting one another. Sometimes a less-efficient system has benefits because keeping the system running is a useful shared project. I don't know if those things are happening today on either site however: both have fairly small communities.

Then the question: is it worth the effort to make the Incubator work as well as beta.WV for its editors? It seems Incubator devs and maintainers would like to see that happen: that's what Incubator was designed for. And at least some recently active editors on beta.wv would be fine with a move. But some of the longer-standing editors there would not: they like it as it is; they helped build it; they don't see any benefits to moving. I think for a merger to make sense we'd need interest from those people to make the move and make Incubator match their workflow. Basically, a few focused people just trying to get a wiki up and running, can do it (technically) anywhere and be happy. But the people who are actively trying to encourage and facilitate new language-development on beta.WV, and the equivalent people on Incubator, should probably talk to one another. SJ talk  17:25, 2 July 2017 (UTC), + 02:11, 11 July 2017 (UTC)

In short, yes. There's no technical limits that (even in theory) holds up this migration, someone asked mw:Extension:Quiz, so we can also install it on Incubator. If they don't want to merge, then let's split Wikiversities out from WMF, because they're violating Founding principles and simply the vision. --Liuxinyu970226 (talk) 02:07, 3 July 2017 (UTC)
Are you talking about the Wikiversity projects currently incubating or about the sister project as a whole? I would think the later would be out of scope of this proposal. If you are talking about the former, could you elaborate in your concerns? Regards, Lsanabria (talk) 15:16, 24 July 2017 (UTC)

What is slowed down by having many small incubators[edit]

Are separate sites actually better & easier for some communities? Should we consider splitting out beta sites for each sister project? From the discussion so far, it seems that many aspects of supporting small communities: from the technical and maintenance notes above, to providing statistics and metadata about project development, are easier on a single site. But it would be good to have examples of maintenance, translation, outreach, communication that work better on a united incubator. And we should definitely improve the quality of community discussion and coordination on the Incubator: people working on new projects have a lot in common -- from dealing with translation, MW details and upgrades, and LangCom to searching for language enthusiasts in small language communities. SJ talk  17:25, 2 July 2017 (UTC)

Unity vs. splittism?[edit]

This is the reason that I oppose "Structured Wikiquote", because that is also a splittism proposal. --Liuxinyu970226 (talk) 02:09, 3 July 2017 (UTC)

No, to the contrary, it is a unification proposal. --Vogone (talk) 18:33, 3 July 2017 (UTC)
I agree with Vogone. Structured WQ, and Unified Wiktionary, are proposals at a lower level: sharing the raw data of [quotes in original language] or [terms, meaninges, and definitions in original language]. On top of that, language communities could still choose to have separate wikis for discussions and front-end presentations in their language; or to share a single multilingual wiki. WikiData is gradually solving the data-layer issue. SJ talk  01:56, 11 July 2017 (UTC)

Why give the Incubator community a load of incubating Wikiversities?[edit]

As said in #Comments, the Incubator is filled with a lot of incubating Wiki sites. The amount of them is very large, including the Wikipedia ones. From zero to two sister projects have been annually graduated from the Incubator since 2014, including one Wikinews this year. Do those at the Incubator have the energy and time improving any Wikivoyage, Wiktionary, or any other incubating non-Wikipedia site? Can those at Incubator invest time and energy on improving one incubating Wikiversty? --George Ho (talk) 08:03, 9 July 2017 (UTC)

The users of Incubator are the users of the test-projects obviously. --MF-W 13:59, 9 July 2017 (UTC)
This seems like a non-issue to me--what would be the burden exactly? Someone who is working on a Ladino dictionary is probably not interested in learning resources in Dutch (altho it's not impossible). —Justin (koavf)TCM 16:42, 9 July 2017 (UTC)
I changed from "burdening" to "giving" in the header, Justin. I hope it sounds better. --George Ho (talk) 16:56, 9 July 2017 (UTC)
Hi @George Ho: -- I see that StevenJ81 has given a thorough and lovely response below. Incubator was designed to be a place to help all sister projects through this earliest phase of development. SJ talk  01:56, 11 July 2017 (UTC)

Clearing up some misunderstandings and misinformation here[edit]

As the administrator arguably most active on Incubator right now—and therefore the individual most likely to be "burdened"—I feel a need to clarify some misinformation that I see in this discussion. For the record, my own opinion on this proposal is neutral.

"Can those on the Incubator invest time and energy on improving one incubating Wikiversity?"[edit]

I truly do not know how things work on Beta Wikiversity. If what happens there is that people working on all Wikiversities in incubation help each other out, then it will be up to all of you to recreate that culture within the Wikiversity tests on Incubator.

  • The way things normally work on Incubator is that each test project—or at least each set of test projects in a single language—has a community of users, and for the most part those different communities of users are independent of each other and focus on their individual tests. Communities can have "test administrators", who have access to most of the tools a regular administrator does, and a community with a test administrator might need very little support outside the community itself.
  • There are two main situations in which tests tend to call on help outside their own communities (before getting very close to approval). (1) They don't have a test administrator, so ask Incubator administrators to perform certain managerial functions (page deletions mostly, and blocks rarely). (2) They need technical help on certain things (like template creation). I did a lot of (1) for Dutch Wikinews, and a lot of (2) from time to time for other test projects, and it really isn't much of a burden.
  • When a project is really getting near being ready for approval, I can and do help more. How I do so, and what I need to do, I will describe in the next section.
Not many [name the project type]s have been approved recently[edit]

I don't really want this to be a finger-pointing exercise, but for the most part the responsibility for this lies with the Language Committee itself. LangCom has gone quiet from time to time, and projects don't get approved when LangCom is not approving them. It is possible that the rest of us could have done more to push them along—I have been doing so recently, in fact. But it's not (mostly) because people at Incubator, or Beta Wikiversity, or Multilingual Wikisource, aren't doing what they're supposed to.

  • If there are really 21 Wikiversity tests ready for approval, you can rest assured that I would start pushing LangCom to deal with them.
There are enormous numbers of projects in incubation[edit]

True. But there are projects in incubation, and then there are projects in incubation. At Incubator, we have about 1,000 test projects that are nominally open, having at least one page of valid content in their test language. I'm working on getting a better fix on this right now, in fact, but I doubt more than about 50–100 are extremely active at any given time (and not all of them are ready for approval). And that is because people working on the test projects don't stay around. Remember that your average contributor to Incubator doesn't speak most of the other languages for projects in incubation, so can't really help much with content outside his/her community. Administrators try to encourage test projects, but there is only so much we can do to keep people around. (Look at the problems English Wikipedia has with this.) And if you want to say that the rules for project approval are too burdensome, and that that is why people leave, then take that up with LangCom, which makes the rules.

Only one Wikibooks project has graduated from Incubator[edit]

@George Ho: I am working on getting a better handle on these 1,000 test projects. But let me give you just a first taste of the problem here:

I have run through all the Wikibooks projects:

  • There are 67 such test projects in Incubator with at least one page of valid content. Of those,
  • ... 27 meet one or both of the following criteria: (a) At least 25 mainspace pages, and/or (b) any material content addition since the beginning of 2016. And of those,
  • ... one (1) currently meets the activity level required for approval: three editors, with ten edits each, for at least three consecutive months. That is Bashkir Wikibooks. It is the only one of the 27 that has met that requirement at any time in the history of Incubator. And it is big enough that LangCom really should be evaluating it for final approval. So I will go to LangCom shortly and push for that.

Additionally, if you consider the true scope of Wikibooks, it should be no surprise that language communities that just barely manage to get a Wikipedia approved are not also doing a lot of writing of original books and manuals—certainly not at a pace to justify a whole Wb project.

If there is something you think Incubator should have been doing to stop projects from gathering dust, please let me know. Creation of approvable wikis first requires a contributing community, and it's not Incubator's job to find that community. StevenJ81 (talk) 18:21, 10 July 2017 (UTC)

"Original research is allowed in Wikiversities, not in Incubators"[edit]

The rules for projects in incubation is that the general rules that apply to project types in their own subdomains apply on Incubator, too. Therefore, if there is no original research allowed in Wikipedias, then there is no original research allowed in Wikipedia tests in Incubator (Wp test code). If there is original research allowed in Wikiversities, then there would be original research allowed in Wikiversity tests (presumed Wv test code). If I make a few mistakes on such things at first, I trust people AGF and give me a chance to learn how to do things correctly.[1]

  • The assumption is that Wikiversity-specific extensions like Quiz will be available to you on Incubator.
  • The question of user flags is something that will have to be worked out in a certain amount of detail. We have only nine sysops and six 'crats on Incubator. But we have over 30 "test administrators"–people who have pretty nearly all of the regular administrator toolkit, and who use those tools within their test communities. I suspect that many Beta WV sysops will end up with that flag, which will allow them to do everything they can do now. I do hope that one or two of you, at least, would be willing to take on the full sysop and 'crat role and help us learn how to take care of Wikiversities, if this comes to past.

To conclude, I think you can make a case either way for this. I personally think that there would be more support for WV tests that are ready for approval if they were part of Incubator than there is now, but I can't say merging would come with no downsides. But for those of you currently in Beta WV, don't assume the worst. I think for the most part we are ready and willing to help you. StevenJ81 (talk) 22:55, 9 July 2017 (UTC)

Is something keeping good small WVs from being approved and founded?[edit]

Marshallsumter wrote above: "Before these proposals to close BetaWV and distribute it to Meta and Incubator, some 14 Wikiversities became open. Since these closure proposals began only 2 became open or have approval ongoing! There are 21 more Wikiversities eligible for approval. The biggest effect of these repeated closure proposals has been to discourage additional eligible Wikiversities from reaching for approval before opening."

This is a conversation worth having. I'm not sure whether it supports or opposes a merge, however. The rate of new wiki creation has dropped across the board, with or without Incubator. But perhaps something - the approval process, the discussions about a move? - have slowed things down.
  • The 21 WV's "eligible for approval" -- this seems to mean projects with interface translation and minimal content, but not the necessary activity. Most don't have recent active editors. Input from people working on those projects would be compelling.
  • Why do you think discussion about a merge has been discouraging? Other than the domain change, none of the proposals has ever suggested this might impact the ability or speed of approvals.
  • Have you seen changes in the approval process slowing things down? As Steven noted above, LangCom makes the rules on approval (and may be open to feedback). And in theory standardizing how things work on Incubator should make it faster to start a project and faster to review + approve it. SJ talk  03:31, 11 July 2017 (UTC)

Another quote: "Apparently, it hasn't occurred to the closure proposers that like the additional language Wikipedias, Wiktionaries, and the many others, most participants also contribute money to the WMF for the pleasure of enjoying each project in their beloved languages and translating among them."

Marshall, please assume good faith :) This proposal is for a merge into Incubator, specifically in the hopes of improving translation, avoiding namespace collision across languages, and the like. The supporters of the proposal believe this would make things easier for new projects to get off the ground. You may be right that beta.WV is superior, and I hope you'll elaborate how. But please don't sneer at the intent of those trying to make Incubator work for everyone. SJ talk  03:09, 11 July 2017 (UTC)
No "sneer" intended! AGF is there, but the chronological evidence strongly suggests that these closure/transfer efforts begun in 2008 have seriously discouraged and distracted participants from bringing more Wikiversities online. I'll take a closer look at the pre-May 2008 effort to see if the enthusiasm supports this exactly. It may not, but if it does perhaps we should consider bringing this to a halt. The first 5-year discussion ended in no consensus! The second 4-year one in rejection! The message seems clear about interfering with or perhaps trying to stop the Wikiversities. Maybe it would be better to consider how to bring the "hopes of improving translation, avoiding namespace collision across languages, and the like." to BetaWV. --Marshallsumter (talk) 03:42, 11 July 2017 (UTC)
The second was rejected on technical grounds, but if you insist, it would be more accurate to say that it also ended in no consensus. StevenJ81 (talk) 14:18, 11 July 2017 (UTC)
  • Okay, here's the chronology before and during the first 5-year discussion: 2006 - 3 wikiversities opened, 2007 - 2, 2008 - before 27 May - 2 (although I'm not sure if the ja.WV was before 27 May), after 27 May - 2 more opened, 2009 - 2, 2010 - 1, 2011 - 1, 2012 - 1, and 2013 - 1, no more until after second proposal rejected, then hi. approved. The statistics are weak but the peak is in 2008 both before and after the start of the first proposal, then downhill during it and zero during the second. I'd say the enthusiasm was severely dampened by the first and all but wiped by the second. Sorry - but however well-intended these efforts here may have been it appears that they severely hurt those contributors who wanted their own WV. 21 eligibles may have been lost, but I hope not. --Marshallsumter (talk) 04:20, 11 July 2017 (UTC)
And in 2017, one Wikiversity is approved, but Koavf said that that project should not created as that's containing email address and machine translated works (cf. [1]). --Liuxinyu970226 (talk) 07:41, 11 July 2017 (UTC)

And... the task is currently stalled for now. --George Ho (talk) 07:44, 11 July 2017 (UTC); struck due to update. --George Ho (talk) 01:37, 10 August 2017 (UTC)

Here's a suggestion: pick 3-5 of the 21 wikiversities not yet approved and let them exist in both betaWV and incubator. Let's see which brings them to approval and open soonest! --Marshallsumter (talk) 04:38, 13 July 2017 (UTC)

@Marshallsumter: Interesting idea, and I've often thought the same thing. But here's the problem. Ultimately, projects are not approved unless there is a community of speakers that works on creating content. And it's just not reasonable to assume that we can duplicate that effort in two places in order to create this competition. (I have to say: if you think Beta WV can help get projects over the finish line without that, then Beta needs to stay separate, because that's not how things work on Incubator.)
  • As a parallel exercise to this, and originally for reasons totally internal to Incubator, I've been working on evaluating the tests we have right now at Incubator. So far, I've gone through the Wikibooks, Wikiquote, Wikinews and Wikivoyage projects, which together constitute about 30% of the tests on Incubator. See the analysis here. Exactly two tests in these groups (out of over 300) have enough activity under current policy to be approvable: Bashkir Wikibooks and Hindi Wikivoyage. Perhaps four or five other tests (out of over 300) either have a current level of activity just a little short of policy requirements, or met policy requirements at some time in the past. No other tests were even close.
So I see the "failure to approve" projects like this to be more a function of a lack of sustained work on the tests rather than a "failure of Incubator". And if the issue is that the activity requirements are too strict, then take that up with LangCom. But LangCom's concern, rightly, is to do as much as possible to make sure these projects will remain active once approved.
Why don't you tell me three Wikiversities that you think are most ready for approval (aside from Hindi) and let me look through them? I'll see what I can do to help you get them over the finish line. StevenJ81 (talk) 12:37, 13 July 2017 (UTC)
  • I'm not sure how much duplication beyond Main Pages, to be merged later, is needed. If the one on BetaWV is say five resources short or five contributors short and three contributions occur on incubator, and approval follows then incubator wins! Same thing with whatever requirements remain. I'll pick five randomly. Hope you can try them! --Marshallsumter (talk) 14:17, 13 July 2017 (UTC)
  • How about Bulgarian (bg.), Haitian (ht.), Mazanderani (mzn.), Polish (pl.), and Tajik (tg.)! I've got two from Europe, maybe Chinese (zh.) as an alternate for one of them. I did not see any from Africa, but if I missed one that might make a good second alternate. Alternates are also in case of a drop out or if you'd rather replace one from Europe. --Marshallsumter (talk) 14:44, 13 July 2017 (UTC)
We're not doing that, and we're not going to run any parallel incubations. I'd like to know what WV tests you think are close to approval and look at them myself. I'd be very happy to get them approved right out of Beta, or to give you my own advice on what needs to happen to get them out. As I said, I don't much care what happens here. But if you think that Wikiversities are stuck for some reason, I'd like to get at the reason why that might be. StevenJ81 (talk) 14:28, 13 July 2017 (UTC)
  • In terms of high page counts Lithuanian (li.) has 327 and Persian (fa.) has 376, if these would be better than Haitian (ht.) or Tajik (tg.), which have low or zero page counts. --Marshallsumter (talk) 14:55, 13 July 2017 (UTC)

Convenience link: States of Wikiversities StevenJ81 (talk) 15:00, 13 July 2017 (UTC)
@Marshallsumter and George Ho:

  • We've got to go with reasonably high page counts for the time being. There are really two issues on the table here, and we need to be careful to keep each in mind, even though they overlap a bit:
    • Could the fact that there is a Haitian Creole community already in place at Incubator help build an HT Wikiversity from essentially zero?
    • Why are fairly developed Wikiversities in Beta not approved?
Personally, I'm far more interested in the second issue. After we start getting a handle on that, we can decide if Beta or Incubator is a better place to try to solve the first issue.
  • Actually, the Chinese Wikiversity (zh.) has the highest page count at 710! --Marshallsumter (talk) 15:28, 13 July 2017 (UTC)
  • I need to point out a couple of technical issues here, which I also mentioned at Talk:States of Wikiversities:
    • Your {{Catanalysis}} template wasn't working right. It was providing edit data only for Category-space pages within the root category of a test, not for mainspace pages. So someone outside your community trying to take a quick look at statistics by clicking a link at States of Wikiversities isn't going to see enough activity to suggest approvability. I fixed that.
    • I noticed that User:Jayprakash12345 has mostly been using वैकल्पिक HI rather than HI to categorize his Wikiversity project. It may be approved anyway, but that kind of thing is a good way for people outside your community looking for numbers not to get an accurate picture. I strongly suggest renaming your root categories back to their basic language codes, so that people looking for approvability numbers won't have a problem finding the information they need. StevenJ81 (talk) 15:13, 13 July 2017 (UTC)
      • ... ... ... I will stick to my position for now, but I might change my mind later then if creating Hindi Wikiversity is delayed for too long. --George Ho (talk) 15:17, 13 July 2017 (UTC); partially struck due to update. --George Ho (talk) 01:37, 10 August 2017 (UTC)

First cut[edit]

I looked at six test projects at Beta (excluding Hindi) having page counts (on States) of at least 200 pages. (I'm not saying 100 isn't enough. I'm just starting with these six.) Understand that I do not speak any of these languages, so I cannot evaluate the actual content, which appears to be part of the issue with Hindi. But here's what I see right off the bat:

  • None of the tests (other than Hindi) currently meets the activity standard (at least three non-bot editors, with at least ten edits per month, going back at least three continuous months)
  • ZH (711 pages) and LT (327 pages): These had stretches—even long stretches—of meeting the standard until fairly recently. If you want to argue that LangCom should have addressed these at the time (2014–2016), and that people left because of a lack of progress, that could be a fair point to make. Still, because of the lack of current community involvement, LangCom will be concerned about whether the wiki will be maintained after approval. But if you can get the people who worked on these wikis before to commit to resuming their work, I'm willing to try to help you make a case to LangCom to evaluate these two projects for approval without waiting for a new three-month window. LangCom will still evaluate these for actual content, and I make no promises on that.
  • ET (261 pages) also had stretches of meeting the standard, but the last month in which that was true was January 2012, over five years ago. So I think LangCom would wonder about current community support.
  • BG (319 pages), FA (376 pages) and UK (296 pages) have never met that standard, even for one month. (OK, I lied: One did, for one month only.) So while the content has been built over time, LangCom would have no assurance whatsoever that these projects could be maintained if approved separately.

It's a different question whether Incubator can provide additional community support for these projects.

  • Since the six projects hosted in Incubator (Wikipedia, Wiktionary, Wikibooks, Wikinews, Wikiquote and Wikivoyage) all already exist as approved projects in Chinese, Persian and Ukranian, I cannot say off the top of my head that there will be a lot of support for native speakers at Incubator. (That having been said, there are many projects in various Chinese dialects, a few in languages close to Persian, and a few in other Slavic languages, so maybe there is more there than I am aware of.)
  • There are projects in Bulgarian, Lithuanian and Estonian. Only one has been extremely active lately, but perhaps the presence of Wikiversity projects also would help. (The one exception is Lithuanian Wikinews, which has had a single, very active IP editor recently.)

Naturally, we can provide administrative support on Incubator, as well as advice and experience getting projects over the top, should the tests move there.
Ultimately, though, nothing is going to get any of these six projects approved without some more sustained community involvement. And while either Beta or Incubator might be somewhat helpful on that front, ultimately a community itself would have to come forward to make that happen. StevenJ81 (talk) 15:57, 13 July 2017 (UTC)

  • Perhaps some form of crossWiki advertising would help! I found the incubator LT Main Page but not yet the betaWV LT Main Page. If these advertised the other, this may help both. --Marshallsumter (talk) 17:40, 13 July 2017 (UTC)
    I'll put notices out on Incubator. I suggest you make the approach on Wikipedia and Wiktionary first, because you understand the needs of Wikiversity better than I do. StevenJ81 (talk) 18:08, 13 July 2017 (UTC)
    Yes check.svg Done on Incubator. StevenJ81 (talk) 18:30, 13 July 2017 (UTC)
  • As a first test, I left the following message on Vogone's talk page: "I have put a sisterprojects display in Lithuanian in my beta:User:Marshallsumter/Sandbox as a way to perhaps stimulate some crosswiki activity! It has the latest addresses I can find and could go on the Lithuanian Main Page if there's interest. What do you think?" --Marshallsumter (talk) 01:29, 14 July 2017 (UTC)
If you're going to work on projects like this, best to learn how to do Wikilinks, rather than URL links. I fixed those.
It's a starting point. But I'd like to see more of a direct call to action on incomplete projects (Wv, Wn, Wy). People need to be told that the projects need active support to be approved. StevenJ81 (talk) 02:59, 14 July 2017 (UTC)
  • Getting the hang of the Wikilinks! Thanks again for your help! Our Main Page gets almost three orders of magnitude more hits than most of our Discuss pages put together. Hopefully, the sisterlinks I created can be the LT version while the other can be EN. That may pull many LTs into those projects including BetaWv and incubator. I liked your calls to action! Usually, I try one-on-one with principal or long-time contributors. Incubator discussion pages such as the ones you chose are good! Our Wv Colloquium page would reach almost everyone that contributes fairly regularly. I'll do some checking around. --Marshallsumter (talk) 23:00, 14 July 2017 (UTC)
  • A one week-long advertising test of our sister projects in Lithuanian has begun! I have recorded 30 d page views for all 15 for the prior 30 d, to see if there's any difference, with local approval and further requests for comments. At the end of the test, I'll record the 30 d numbers again. --Marshallsumter (talk) 20:44, 26 July 2017 (UTC)
Are you sure that lt pages you mentioned are not machine translation? I would ask @Hugo.arg: (the only lt user which I somewhat familiar) to verify. --Liuxinyu970226 (talk) 01:44, 27 July 2017 (UTC)
Hm... What should I verify exactly? I did not completely understand this discussion anyway. Why you want to create/revive lt.wikiversity which never was anything serious (the only user who tried to promote it was cs:User:Kusurija himself being native Czech-speaker)? Currently, lt.wikipedia itself struggles to get new contributors and it is being kept active by small group of long-time users, lt.wiktionary, lt.wikibooks and lt.wikisources are dead, lt.wikiquotes is being kept active by only one user. I think, in this case, we should concentrate creating powers mostly to lt.wikipedia, which is main wiki source for internet users, rather to obscurious side projects which will never be anything impacting. Hugo.arg (talk) 09:36, 27 July 2017 (UTC)
The week-long test was stopped by one of our bureaucrats who disagreed after the test was started but not before and instituted his own solution. Regarding the lt, I used words and phrases as they appear on other lt sites such as Vikipedija. --Marshallsumter (talk) 23:15, 28 July 2017 (UTC)
  • The apparent effects are only dramatic for the smaller lt projects. Nine of the 14 showed sharp spikes upward on the 26th, the first test day. Each was much more than my one visit to go to the Page information. The larger ones: lt.wikipedia, lt.wiktionary, lt.wikiquotes, lt.wikibooks, and lt.wikisource, showed only slight upward trends. The most dramatic gain was to the lt.Incubator (9.3 %) upwards from 2.3 %, second highest was Betawikiversity (3.9 %) off a previous -0.8 % trend. Although the statistics are far short of what was needed, it does appear that advertising on en.Wikiversity was beneficial! --Marshallsumter (talk) 00:08, 29 July 2017 (UTC)
  • The test for the 27th was cut off at 14:55 UTC. The lt.wikiversity jumped upward by 29 %! The lt.incubator was up by 13 %! Betawikiversity was up 3.6 %! The five major lt projects continued as before or were flat to slightly downward. --Marshallsumter (talk) 00:59, 29 July 2017 (UTC)
  • The results for the 28th are mixed. The major lt projects: lt.wikipedia, lt.wiktionary, lt.wikiquotes, lt.wikibooks, and lt.wikisource, showed no real change for individual days on the 30 day graphs. The lt.wikiversity showed a sharp spike downward in contrast to the 26th and 27th. Lt.wikivoyage continued its sharp upward spike as did lt.wikispecies and lt.incubator main page. Lt.Mediawiki showed a sharp downward spike as did Betawikiversity and lt.wikinews. --Marshallsumter (talk) 01:20, 30 July 2017 (UTC)
  • Perhaps the more significant effects of the test are
  1. the presence of the Betawikiversity logo on en. wikiversity Main Page correlated with the test suggesting that adding the project logo to our Sister projects section helps and will help the fledgling wikiversities,
  2. adding the Incubator logo may have little effect generally for the project but may encourage individual projects, and
  3. the current list where lt.wikiversity is included in Other does not benefit those wikiversities in this category. --Marshallsumter (talk) 02:14, 30 July 2017 (UTC)

Hindi Wikiversity[edit]

The Phabricator task on Hindi Wikiversity is closed as "resolved", but we are still awaiting transwiki process of the test site to be completed. Nearly two months after task was created, we have a newer Wikiversity... yet incomplete. --George Ho (talk) 01:37, 10 August 2017 (UTC)

Never mind. The task is reopened. --George Ho (talk) 04:50, 10 August 2017 (UTC)

Gosh. The wiki exists and is being edited by the community. Technicalities always need to be sorted out when a new wiki is created; there is no need to spam this page, which is not related to it, continuously with updates about that. --MF-W 15:23, 10 August 2017 (UTC)
Don't worry. I wasn't spamming, but I'll post just one more update about hi.wv. I struck out most of the pre-update post. I saw the pages transwiki-ed into the official hi.wv. This reaffirms my "oppose" vote. Two months isn't that long enough to change my position about this proposal. I still believe in (the powers of) Beta Wikiversity, despite what others think about it (unless the majority becomes overwhelming to handle). I heard that there were issues with Beta versions before transwiki-ing the pages, especially hi.wv, but I hope that's sorted out without merging Beta WV into Incubator. --George Ho (talk) 18:34, 10 August 2017 (UTC)

Footnotes[edit]

  1. Concerning fair-use images: Regardless of whether Wikipedias and Wikiversities "may" have fair-use images, it is unquestionably true that per WMF rules a project must have an Exemption Doctrine Policy in place for that to happen. The only Wikiversity that has one right now is English Wikiversity. Consequently, the image files currently on Beta WV are technically outside policy. Incubator also does not have an EDP. Therefore, we would not allow Beta WV's current images to be imported. If Beta WV, or an individual test wiki within Beta WV, creates an EDP, then we can discuss that further.