Wikipedia talk:Stable versions proposal/Archive 1

first talk

At what point can any article be trusted to be "completed" and thus publication-worthy? Who then would we trust to update it? violet/riga (t) 14:10, 26 November 2005 (UTC)

it has to adhere to publication standards like how britannica or encarta would do. plus it can always be republished. -- Zondor 14:14, 26 November 2005 (UTC)

It's not clear: is this the same idea as having a "stable" vs "development" version of an article, or are you instead arguing that once a page becomes published, no version is then available for editing on Wikipedia? — Matt Crypto 15:06, 26 November 2005 (UTC)

I am proposing the former: stable vs development. stable or published is not editable. development or normal articles are. -- Zondor 15:10, 26 November 2005 (UTC)

Conditional Support

I like the idea of having articles reach "publication" status; it could be considered the end of the line for article development. However, I do not entirely agree to the concept of page protection, as it is the concept of Wikipedia to quickly fix something when something changes. For example, if all the apples of the world were to spontaneously turn purple, we can be the first to say so. However, you have a point. My idea is that a subpage of the same article could be made, where people can freely edit the content. Whenever it's important to update the main article (like my purple apple example above), we can move over the content to the main article. —MESSEDROCKER (talk) 18:05, 26 November 2005 (UTC)

so yes, it means you do have full support. -- Zondor 18:33, 26 November 2005 (UTC)

If a published article would have no editable version, then I strongly oppose this. Though something as described by Messedrocker above is more workable, I'd like to see it implemented on a test project - or at least described clearly.--Piotr Konieczny aka Prokonsul Piotrus ^Talk 23:46, 27 November 2005 (UTC)

"Publication" -- misnomer?

I think the word "publication" is a bit of a misnomer, as, at least in the scheme proposed, "published" articles would be available on Wikipedia just the same as "unpublished articles". Both types are published on the web, and not in print. — Matt Crypto 18:59, 26 November 2005 (UTC)

being published does not mean it has to be in print. both can be in print also. yes, both are on the web but they can have different status. published vs. normal = checked+protected vs. editable -- Zondor 19:15, 26 November 2005 (UTC)

Yes, I know, but all pages on Wikipedia are "published" on the Web, whether they're new stubs or Featured Articles. I'm just saying that the word "published" is a poor term to differentiate between a "stable" and "development" branch of an article, because all Wikipedia articles are published in the online sense, and most are unpublished in the hardcopy sense. — Matt Crypto 19:46, 26 November 2005 (UTC)

DIY

Why not do it yourself? Make your own encyclopaedia website, made up of Wikipedia articles that you think are good enough. I believe the information is free to copy, right? If we protect a page then the article is never going to improve, things change with time. Even Britannica has mistakes! Gerard Foley 01:19, 27 November 2005 (UTC)

We don't have to divide. Strength in unity. -- Zondor 06:50, 27 November 2005 (UTC) -- Zondor 06:50, 27 November 2005 (UTC)

See Wikipedia:WikiReader. Allready being done. --Stbalbach 23:13, 27 November 2005 (UTC)

Emphatic no!!!!! Lock pages? Then Wikipedia would have failed. What a horrible idea! User:Zoe|^(talk) 03:03, 27 November 2005 (UTC)

Never going to improve? Did you fully read the project page? The proposed system is an extension of the existing system and will not compromise it. The locked pages are on the side, a different namespace called "Published" or "Publication". If the new system is in place, you would hardly feel anything has been different, as all articles are continued to be freely editable. -- Zondor 06:30, 27 November 2005 (UTC)

Any user reading the published version instead of the active version will be 1 (or probably 0.5 or less, but still something) less user to see (and therfore have a chance of editing) the current version, so to the degree this suggestion is a success and widely used, it will slow down wikipedia's growth. I therefore stronly oppose this idea. Amaurea 22:23, 22 December 2005 (UTC)

Over my dead body. It's a joke to go asking if the authors of articles want them to be "published" when there's not the vaguest sign of consensus to go ahead with this at all. Ambi 22:53, 27 November 2005 (UTC)

no one has claimed there is full consensus. this is only a stage in gathering one. -- Zondor 23:07, 27 November 2005 (UTC)

I did read the project page. Never was probably the wrong word, but it will make it harder to fix small mistakes. I like the idea of a reliable version of Wikipedia, but this is not it. Gerard Foley 17:06, 28 November 2005 (UTC)

I don't know what you're all excited about. Please don't have a knee-jerk reaction or take it out of context when someone mentions protection or locked pages. There is an editable version - it is there. -- Zondor 17:57, 28 November 2005 (UTC) look at it this way, the protected pages being introduced is not a wiki whatsoever so you don't need to be so cliched and stamp it Meta:Anti-wiki. let's call it something else like static pages for publishing wiki pages -- Zondor 02:17, 29 November 2005 (UTC)

No I actually didn't read the proposal fully, I was thinking of something else (I don't know what). I was also a bit drunk when I wrote the original response. Talk of this has been going on for years, see Wikipedia:Pushing to 1.0 started by Jimbo Wales 2 years ago. Gerard Foley 06:03, 29 November 2005 (UTC)

Regarding the fact that it is harder to fix small mistakes, it would be harder because publications are a different beast to wikis. Wikipedia's and yours would be put on the on the line. Published versions are thus favourable because it has benefits of being reliable. -- Zondor 02:31, 30 November 2005 (UTC)

Namespace

There could even be a wikipedia.org/published/ as opposed to wikipedia.org/wiki/. - Fredrik | tc 12:37, 27 November 2005 (UTC)

that's a good one. the way we are editing articles freely now should be the main centre of attention. i expect people to keep continually publishing their freely editable wiki article as often as possible. this way we maintain our wiki openess, yet we can be traditional that people are comfortable with. this environment would be very favourable to those who want quality and accurate information like serious researchers and schools. indeed, using the published namespace would be a little ugly. however, it would be great if they can still interlink with each other. -- Zondor 20:34, 27 November 2005 (UTC) it would make sense to move it to /published instead of remaining at /wiki because if they are all protected, they wouldn't be wiki anymore. -- Zondor 20:48, 27 November 2005 (UTC)

Who's publishing?

All these ideas are interesting, but who's doing the publishing, what kind of distribution will it have? There are already Reader projects, in which individuals on Wikipedia publish articles. This page "looks" like somthing official (ie. there is money being spent by the wiki non-profit), but is there accutual official monetary support behind it, or is this just a volunter "project"? --Stbalbach 23:11, 27 November 2005 (UTC)

The name of this page is very unhelpful: this is a proposal for a "stable" vs "development" version of articles, and not a plan for print publishing. — Matt Crypto 23:21, 27 November 2005 (UTC)

All wikipedians would be doing the publishing by consensus. An official publishing house could be set up by the board. Nothing needs to be printed or spend too much money yet because they can all be online (electronic publishing). It isn't official and has potential to be, however, it is closely tied with Wikipedia:1.0 which is more official as initiated by Jimbo. It is still gathering consensus being a {{proposal}} and all. Its a little different with this proposal because articles can be published one by one starting soon. -- Zondor 23:22, 27 November 2005 (UTC)

Needs dedicated software support

I think this proposal would require new features in the MediaWiki software. A new namespace is unlikely to be sufficient. Let me describe what I have in mind, and I hope you'll see why this would be awkward to implement on top of the existing infrastructure.

Suppose Wikipedia was divided into two zones: the published/frozen/1.0 etc. zone, and the editable/hot/development zone (the latter corresponding to the current system).

If you're in the frozen zone and you follow a wiki link, you should be taken to another frozen zone article if one exist, and if there is none yet, you should be taken to the corresponding development zone article. I.e., some sort of fall-back mechanism should be in place.(This can only be done in the current system by keeping an explicit inventory of published articles and updating links that point to published/frozen pages.)

The two zones should be visually distinct, perhaps by using different style sheets.

Arguably the frozen zone wouldn't need talk pages or user pages, and consist only of a subset of the namespaces (main/article, image, portal, category). Categories would be problematic, because they are created dynamically. If a given frozen category contains only articles from the frozen zone, it might end up looking quite empty. But if categories are shared among the two zones, this could easily lead to confusion and/or inconsistencies (e.g. what if a frozen article is in category C and its corresponding development version is not?).

The main reason why talk pages shouldn't exist in the frozen zone is for everyone's sanity: we don't want to treat the frozen version as a branch/fork with separate development. (It's doable in theory, but can be quite messy and confusing in practice.)

Each page in the frozen zone should point to the corresponding development version of the article, perhaps even to the exact revision that got frozen/published. This would require different instructions, and "edit this page" would get a bit more complicated.

I think we should work out a few concrete use cases for this proposal and then ask for developer assistance. To a large extent it's a technical challenge. --MarkSweep (call me collect) 01:22, 28 November 2005 (UTC)

the exact needs of the dedicated software support needs more thought. however, a precursor system like the first system using namespaces can used initially to put this into practice to confirm the usability. -- Zondor 03:59, 28 November 2005 (UTC)

So how should links inside the "Published:" (or "Stable:") namespace be treated? I think the most consistent would be to allow only links to other "Published:" pages, meaning that this namespace will be at first very incomplete. But if you can get enough people on board and manage to get an approval procedure that is both fairly strict and approves 100 pages/day, this problem will eventually go away. (Published:redlinks could also provide a disclaimer and link to standard wiki page (as a later software feature), but should not link directly). Actually I think the setup for the approval procedure and getting enough people here is the hardest part of this project. Kusma (talk) 05:13, 28 November 2005 (UTC)

The other problem is that a separate namespace will essentially just create a fork of existing articles. Those articles would have to be protected, and their talk pages redirected to the development version's talk page. This is rather tedious work which should be automated. Next, Wikipedia isn't just articles. What about images, templates, and categories? The frozen/development distinction is really orthogonal to the namespace issue. --MarkSweep (call me collect) 05:54, 28 November 2005 (UTC)

dont protect, use permanent link to cite

I dont think its necessary to do this, as information keeps changing. For instance if you protected say the article on United States of America how do you update it when there is a new president etc.? If you want something from which to cite, use the Permanent Link over there at the bottom of the lefthand column which gives the link to each version of a page. Maybe a list of permanent links which are seen as the 'best' versions of a page would work? Astrokey44 02:40, 28 November 2005 (UTC)

That is similar to Wikipedia:Baseline revision. Citing a particular version at any given time is not good enough as any other versions in time. This proposal involves a big fussy quality assurance check. Plus, republising as often as possible is encouraged. yes, we can do away with protecting and go with citing oldid's as long as there is an explict quality check. though using a cite link is not so proper, no? -- Zondor 03:41, 28 November 2005 (UTC)

How about some sort of list of articles the moment they became featured, so you could have a page that said something like "this lists wikipedia's pages on the revision date when they became a featured article" - something like: Tom Brinkman became a featured article on 10 August 2005, this is the revision showing when it was featured: [1] , maybe move them to a different link something like wikipedia:featured revision/Tom Brinkman, so that it stands out from just any old version of an article? Astrokey44 05:55, 28 November 2005 (UTC)

Yes, this would be a good idea. You would want to provide a notification on the normal article to point to the featured article. eg. In Tom Brinkman it should say For the published featured article, see Wikipedia:Featured revision/Tom Brinkman. And do you mind that Wikipedia:Featured revision/Tom Brinkman should be protected so that it will remain the exact version? When using the particular oldid, there should be consensus that this is a publication worthy version that is good quality and accurate comparable to Britannica and Encarta. However, does "featured" mean its publication worthy? I chose the word "Publication", though loaded but a strong word to indicate so. Wikipedia:Featured revision/Tom Brinkman looks a little ugly. Why not put it as Publication:Tom Brinkman? Or even better: http://en.wikipedia.org/publication/Tom_Brinkman ? -- Zondor 06:36, 28 November 2005 (UTC)

Yep. Locking or moving articles won't fly. The marker idea (a given version is acclaimed as the "published" version) is probably more workable, but that requires software. Presumably the "published" version is in the version history; version URLs are stable, so we just need a link to that somewhere - David Gerard 14:14, 28 November 2005 (UTC)

Typically, when you go through the version history, a particular version, considered as marked for example, will not necessarily reflect that old version because the latest version of images and templates would be used. unless the mediawiki software needs to accomodate this by storing the oldid of the image or template or store the total raw data of that particular version. -- Zondor 15:30, 28 November 2005 (UTC) plus there can be complications with templates, categories or images being deleted or maybe when moved. -- Zondor 15:40, 28 November 2005 (UTC)

Templates are also a problem for the article validation feature. I've put a suggestion on m:Article validation possible problems that old versions also try to pull in the relevant version of the template, or be stored with the version numbers of any templates. I predict that will be far, far more acceptable than making any article in the main Wikipedia non-editable - David Gerard 15:36, 28 November 2005 (UTC) from [2] -- Zondor 17:14, 28 November 2005 (UTC)

So what you're saying is that the article needs to be protected (under the "Published" namespace or whatever) and the relevant templates and images need to be as well? That makes for a more difficult job. violet/riga (t) 21:24, 28 November 2005 (UTC)

this is concerning another "publishing" system or Meta:Article validation. -- Zondor 02:00, 29 November 2005 (UTC)

Name

I'm not so sure of I like the name "publication" for what is essentially a "frozen" version. However, I like the idea. Some version of an article could be declared "good" and then be moved to the protected "frozen" version, while the standard version continues to change and be updated. The "frozen" version should only be updated every couple of months, or when something important happens (e.g. government changes), and changes from the standard page be incorporated. Note that an endorsement of specific version similar to a freeze already seems to exist for spoken articles (see Wikipedia:WikiProject Spoken Wikipedia). It might be useful to ask there how the version to be recorded is chosen. Kusma (talk) 02:41, 28 November 2005 (UTC)

"Publication" implies quality and protection. "Frozen" implies protection only. Britannica and Encarta have "Publication". Wikipedia only has "Frozen". Schools prefer "Publication" over "Frozen" which means schools prefer Britannica and Encarta over Wikipedia. -- Zondor 03:46, 28 November 2005 (UTC)

Oh, I see. You are right and the name should definitely not be "Frozen" since that doesn't sound good from an outside point of view. I agree that "Publication" is superior (not sure if optimal). Kusma (talk) 03:53, 28 November 2005 (UTC)

But by designation one version as "publication", this creates the wrong impression that other articles somehow haven't been "published" yet. What we really need is a mechanism that's slightly more conservative than editing a page and having one's changes be immediately visible to the whole world. Ideally, the visible version of a page (the one visible to everyone when they look up an article) would lag behind the most recent revision just enough to discourage vandals, but not enough to frustrate serious contributors. I'll predict that we'll see a drop in silly vandalism if edits don't have an immediate effect but require some amount of additional review. --MarkSweep (call me collect) 06:03, 28 November 2005 (UTC)

the impression is right - articles that have not been published does not deserve the title - it has not been through quality assurance. -- Zondor 06:16, 28 November 2005 (UTC)

Zondor, "published" doesn't mean high quality. It means "to prepare and issue for public distribution or sale". All content on Wikipedia is published, by definition. — Matt Crypto 15:54, 28 November 2005 (UTC)

"Publish" can serve another meaning that we want to tell the world we recommend you to use this stable version because it has been quality checked rather than the one we are working on at the moment. -- Zondor 02:02, 29 November 2005 (UTC)

It's not a wise idea to take a word that's generally understood to have one meaning, and reuse it with a different intended meaning. It'll only confuse people. Alternatives include the "stable" vs "development" terminology. — Matt Crypto 10:07, 29 November 2005 (UTC)

What Matt said. --MarkSweep (call me collect) 10:29, 29 November 2005 (UTC)

That sounds ok as long as editing continues on the 'normal' article - but how would you determine who could edit the 'frozen' version, and when? I suppose it would mean admin-only editing of the frozen version, although there might be alot of people unhappy about that Astrokey44 05:58, 28 November 2005 (UTC)

I would like a proposal such as "Editing continues normally on the standard wiki page, but every now and then a good version is decided upon (at most once per month) and moved to the "Published:" namespace by an admin." I think administrators should not edit these pages, just update them by replacing with a new standard wikipedia version when decided so at Wikipedia:Requests for Publication/Updates. Kusma (talk) 06:21, 28 November 2005 (UTC)

Sounds like a good idea to me. --JK the unwise 15:39, 28 November 2005 (UTC)

Do we really want this?

IMHO, there's no need for any drastic measures to accomplish what you want, especially not freezing pages. All you need to do is to make a template which basically says "This article has a version we think is good, click here to see it." which would appear on those articles that have them. Watching the changes to the tag will be as hard as watching any other category, but a list of links to good versions and a bot should probably manage.

But the question is, do we really want this at this point? Who will decide when it's good enough? A committee of experts appointed in some way? Do we have the required infrastructure for checking credentials? Or do we just duplicate FA procedures and go on to publish articles which we claim can be held up to standards of professionaly reviewed encyclopedias and hope for the best? If we get just a few wrong, we'll be ridiculed for years. It's not that they don't make mistakes, but we're still the underdog here and more vulnerable to damage from gaffes. Zocky 16:10, 28 November 2005 (UTC)

Yes, we do actually. Nobody is saying that we freeze the main articles, no indeedy. What we have been (rightly) critized for is instability and inaccuracy. Having a page frozen in its own namespace is a good idea. If problems occur, then we can add a tag saying that we don't want to use this as the frozen page anymore and then we can move on to the next good revision. - Ta bu shi da yu 10:49, 29 November 2005 (UTC)

The main version of an article is the one most people read, not the one we declare to be the "main version". If most people read a frozen version of an article, essentially a non-wiki version of the article, then those people probably won't participate in the wiki process. What makes a wiki special is that everybody can change the pages. If we change that into "everybody can change some behind-the-scenes article which most people don't really bother to look at", then we will not only lose most edits (the ones from average readers), but also our main avenue for recruiting new editors. Wikipedia has grown this quickly this large, and gotten the astounding amount of quality material is has precisely because of the way it has worked up to now. Using static versions of articles sounds like an excellent way for wikipedia to suffocate itself. Amaurea 22:38, 22 December 2005 (UTC)

Do we really want this at this point?

It is certainly needed if Wikipedia wishes to compete with Britannica and Encarta in quality and accurateness. Wikipedia is built quickly without the burden of accuracy. It serves a different market compared with the traditional encyclopaedias who target schools. If the community does not want be known as a reliable source then so be it.

Who will decide when it's good enough?

Wikipedia:Consensus

A committee of experts appointed in some way? Do we have the required infrastructure for checking credentials?

The level of content, quality and accurateness is a function of its Wikipedians. If we have expert Wikipedians in this area then we will have good articles in the same area. That's the best we hope to do unless we hire some experts.

Or do we just duplicate FA procedures and go on to publish articles which we claim can be held up to standards of professionaly reviewed encyclopedias and hope for the best?

The featured article process is very similar to Requests for Publication and may come together to some arrangement.

If we get just a few wrong, we'll be ridiculed for years.

Wikipedia is full of wrongs or pretend it does not have any. This is why Britannica and Encarta are favoured.

-- Zondor 16:44, 28 November 2005 (UTC)

I dont think we have to beat Britannica and Encarta at everything. Lets just do what we do best and leave them to do what they do best. We're not trying to take over the world here --- Astrokey44 00:59, 29 November 2005 (UTC)

Microsoft Encarta has already started trying to beat us: noticed their new "edit this article" feature? -- Zondor 01:53, 3 December 2005 (UTC)

This is a shockingly bad idea. Trollderella 20:48, 28 November 2005 (UTC)

Care to elaborate? - Ta bu shi da yu 12:24, 6 December 2005 (UTC)

How about tagging a stable revision with a link on the page to that revision. e.g. "This revision has been certified by X" ? The main article page is still editable but there is a revision that people can look at if they feel the need to read a certified version? Or am I bringing my experience with software development into an entirely different field? - FrancisTyers 16:53, 30 December 2005 (UTC)

Please stop spamming talk pages

While my initial opinion was that this is a good proposal, if people keep spamming talk pages with requests to "publish", especially talk pages of non-featured articles and while this proposal is still being discussed, I for one will turn into an ardent opponent. -- Jmabel | Talk 17:58, 28 November 2005 (UTC)

Non featured articles are acceptable. Namely, Wikipedia:Good articles or Wikipedia:Standard articles which are v0.5 or usable as described in Wikipedia:1.0. -- Zondor 18:00, 28 November 2005 (UTC)

That reply really missed the point. I applaud your efforts to get your proposal off the ground, but spamming various talk pages is not the correct way to go about things. violet/riga (t) 21:19, 28 November 2005 (UTC)

Does FA work, anyway?

Before becoming too enthusiastic about how well the Featured Article program works, maybe we should review a sample of the articles and see whether they were really all up to scratch.

For example, list of countries with nuclear weapons became a featured list while listing Australia as a state formerly possessing nuclear weapons or programs. The entry describing Australia was also inaccurate on several details, all of them apparently from the same political source, see Talk:List of countries with nuclear weapons. The current entry is still misleading. Australia has never had a weapons program! Andrewa 18:10, 28 November 2005 (UTC)

In addition many FA's became so before the standards were higher and have remained part of the FA cannon but would never pass if nominated today. Granted there are steps to de-list a FA, but it's not fun or easy to do, many weak articles remain. --Stbalbach 18:29, 28 November 2005 (UTC)

Support

I support this effort, but appearently for reasons not stated on the project page: I'm hoping it'll discourage vandals and POV pushers. Mathematics is perhaps a poor example: it has a lot of peple watching it, and has a lively talk page; objectionable edits are quickly reverted. A more appropriate example might be an obscure example, such as Maslow's hierarchy of needs, which is an adequate article that has seen nothing but frequent vandalism since approximtely forever. As its talk page will attest, at least one editor left on account of this.

At least part of my interest in publication would be eliminated if WP had more reliable, more robust ways in which a group of editors could share the burden of reviewing watchlists. For example, some very fast/simple collective voting mechanism by which a regular pool of editors could note that they reviewed & agree with any given recent article change. Since "publication" would be governed by some voting process, some discusion of streamlined voting proceedures might not be inappropriate.

But perhaps this is the wrong forum in which to discuss these issues. linas 22:37, 28 November 2005 (UTC)

yes, this is a good reason. i am not sure it will alone discourage vandals but it can certainly help hedge against the risks of vandalism. -- Zondor 02:08, 29 November 2005 (UTC)

It will certainly help. The John Seigenthaler Sr would be a good place to start, methinks. - Ta bu shi da yu 02:40, 2 December 2005 (UTC)

A related proposal

Those of us who edit the George W. Bush article have been frustrated about the high rate of vandalism. It's probably the most frequently vandalized page on Wikipedia. From time to time some admin reacts to the problem by protecting the article, which I personally consider a bad move.

Aside from the nuisance factor, there's a concern that many readers who visit that article will access it while it's in a vandalized state. One possible solution presented at Wikipedia:Experimental vandalism protection and discussed on Wikipedia talk:Experimental vandalism protection is similar to the "publication" idea but much more limited. Once a week or so, an admin who never edits the article for substance would identify the most recent unvandalized version and save it as "George W. Bush - scrutinized" or some such. The main article would remain open for editing by anyone, even anons, just as it is now. The difference would be that it would have a note at the top mentioning that it's frequently vandalized, and that a recent nonvandalized version can be found here (wikilinking to the "scrutinized" version, which would be protected).

I favor starting with a "scrutinized" version only of George W. Bush, a highly visible article that's often vandalized. We might consider expanding it in the future.

Unlike the "publication" proposal, establishing a "scrutinized" version wouldn't address article quality issues other than vandalism. I mention it here because some people interested in Zondor's proposal might also be interested in the other one. JamesMLane 02:28, 29 November 2005 (UTC)

just a thought: hopefully without increasing Meta:Instruction creep, there could be a marker system that indicates which was the most recent scrutinised version (a little similar to Meta:Article validation). only one single marker per article. only those with high level counts or admins can change that marker. or maybe even multiple markers per article to indicate multile scrutinised versions. this wouldn't cost too much to change the mediawiki software. -- Zondor 03:14, 29 November 2005 (UTC)

One advantage to having a separate "scrutinized" version, instead of just a marker, is that, if the system seemed to be working, wikilinks in other articles could be changed to piped links, in the form [[George W. Bush - scrutinized|George W. Bush]]. That way, readers following a link would automatically reach an unvandalized version. Of course, a note on the scrutinized version would direct the reader to the editable version. JamesMLane 08:59, 29 November 2005 (UTC)

Previous marker idea: Meta:Article marker feature -- Zondor 17:38, 1 January 2006 (UTC)

Renamed

I've renamed this proposal "Stable versions", rather than "Requests for publication". As argued above, the "publication" terminology is simply incorrect. A) this proposal isn't particularly about any new form of publishing (e.g. print), and B) all Wikipedia articles are published on the Web, and this proposal doesn't change that. I suggest the "stable" versus "development" terminology — but there could be better. — Matt Crypto 12:02, 29 November 2005 (UTC)

I agree with the reasoning for the move. violet/riga (t) 12:03, 29 November 2005 (UTC)

 

"Stable" does not imply quality and accurateness that you would get in a publication. Please find a word that does. Maybe Released version? [3] [4] Names are powerful and magical and should be chosen carefully. People would still think: "Stable? I don't want stable. I want quality and accuracy. I want a published version that you would get from Britannica and Encarta". -- Zondor 02:02, 30 November 2005 (UTC) Revised version? -- Zondor 02:11, 30 November 2005 (UTC) 1.0 versions/editions? -- Zondor 03:04, 30 November 2005 (UTC)

You say that "Stable does not imply quality and accurateness that you would get in a publication". I think you have a false association of publication with quality. You can publish things of poor quality in any medium, and, of course, there are many poor Wikipedia articles that are published on the Web. So let's reassert that "published" is not a synonym for "quality" or "accuracy". Think of "published" as meaning "made public". — Matt Crypto 08:13, 30 November 2005 (UTC)

Exactly. There is no inherent connection between published content and quality content. In scientific publishing, though, it's common for things to only get published after they've gone some form of peer review (and one can debate the merits, flaws, and cost of that system, but this is not the place). We do have peer review, but I don't think "stable" should necessarily mean having undergone peer review, if only for the fact that peer review is often slow. In fact, the peer review process on WP exists partly to solicit input on articles that don't have an active body of editors that collaborated in its creation. I'd guess that there is a vast number of articles out there which can be considered stable, aside from the occasional vandalism. If an article has a group of regular editors that patrol and update it, the decision of whether it's sufficiently stable and which version to pick for freezing can be pretty much left to those editors. --MarkSweep (call me collect) 12:10, 30 November 2005 (UTC)

Interest in the project

I know this is at a very early stage, but I think there would need to be a massive amount of interest from the community in order for this to work properly. I see it as a poor relation to FA status, and know how the FAC process can sometimes draw too little attention. The intention is good, but I personally don't think that this solution will work in practical terms. violet/riga (t) 12:09, 29 November 2005 (UTC)

I think it can, but it needs to be thought through a lot better as to how it would actually work; if it were to proceed along current lines, it would very much be poor relation to FA status. Ambi 03:52, 30 November 2005 (UTC)

I agree with violet/riga and Ambi, except that I think a "poor relation to FA status" is a reasonable goal. It would involve a much lower standard than what Zondor has proposed -- something like "no factual inaccuracies, no major omissions, no misspellings or grammatical errors, no vandalisms, and no NPOV disputes". Instead of aiming for a "stable" version, it would be considered merely a "scrutinized" or "reviewed" version. It wouldn't necessarily be stable, because the reviewed version could be updated from time to time by the copying of the main article (which would remain available for open editing and which would, we hope, continue to improve). Obviously, this wouldn't accomplish all of Zondor's objectives. It would, however, require significantly less work. JamesMLane 16:12, 30 November 2005 (UTC)

But that's the point. How do we know there are no factual inaccuracies? How do we know there are no major omissions? FAC does a reasonable job of achieving this, but by no means a complete one. If this is to be any different to FAC, it really does need to come to some means of solving those first two issues. Even more useful, however, would be to do this with Wikipedia 1.0 in mind, and ask two additional questions: is this important enough to go in a print distribution, and is this in the appropriate form to go in a print distribution? Ambi 16:31, 30 November 2005 (UTC)

Given enough eyeballs, all bugs are shallow. It is best if the total combined effort, every bit of effort we have in Wikipedia (Wikipedia:Good articles, Wikipedia:Standard articles, Wikipedia:Featured articles, and others) is efficiently directed towards creating the end product, a published/stable article version. I think this proposal overlaps them or they are a subset of this proposal. There is an issue of articles achieving a status only then to be degraded over time - this proposal can prevent that. -- Zondor 04:33, 1 December 2005 (UTC)

Sources, for one. Lots and lots of sources, each one of them verified. See MDAC for an example. I know it's a lot of work, but that's what it would take! - Ta bu shi da yu 02:43, 2 December 2005 (UTC)

New language

perhaps its best to put these published/stable versions on a new language site (and be protected?) in order to work seamlessly with existing languages. -- Zondor 14:16, 30 November 2005 (UTC) -- Zondor 17:17, 30 November 2005 (UTC)

Now you're getting towards the realms of a total Wikipedia fork. violet/riga (t) 16:46, 30 November 2005 (UTC)

it also conflicts with people's reaction to protection in wikipedia. at least a new haven can be created dedicated to protection. -- Zondor 03:24, 3 December 2005 (UTC) It shall be called Nupedia. -- Zondor 04:15, 3 December 2005 (UTC)

Scaling

I think that maintaining a stable wikipedia is a worthwhile goal, but I don't think that this process would really scale to the job. If we lowered the standards to a version of an article that has no vandalism, is referenced and someone has done a quick check of its references and allow this to stand as the stable version. This could then be compared with the 'live' version every month/week/day, depending on how busy the article is, and someone (admins, regsitered users, somewhere in between?) could update the stable version accordingly and set this as the new live version. I think this would get through a lot more work than putting the whole thing to a committee and seems to perform more or less the same function. Comments, anyone? --Cherry blossom tree 23:51, 30 November 2005 (UTC)

It'd be a lot of work for what, I think, could be questionable benefit; the amount of articles affected by vandalism at any one time is negligible, and any checking of references is going to slow the process right down, in which case we may as well do it properly. Ambi 00:32, 1 December 2005 (UTC)

I strongly disagree with the comments in one respect: vandalism is a large concern, even if only some of them are actually vandalised. For instance, the John Seigenthaler Sr was vandalised, and caused us a great deal of credibility problems for us. I think now is the time to start thinking of ways to show that we have more reliable and stable versions of articles. One thing is to try to remember that we're writing an encyclopedia here! - Ta bu shi da yu 02:45, 2 December 2005 (UTC)

Yes, it's true that checking (off-line) references would be time consuming. And I suppose anything contentious would be ironed out before then. I'd be prepared to let that drop. But if we did this by review board, even going many times faster than WP:FAC goes we'd still only have a couple of thousand in a year. I was just trying to propose something that might get through the pile a bit quicker. --Cherry blossom tree 00:44, 1 December 2005 (UTC)

Not necessarily. If this was done with Wikipedia 1.0 in mind, it would not be feasible to have copies of all of every article; rather, it would just be a summary short enough to be printed. I don't think it would be nearly as hard to impose strict standards of referencing and quality on articles that really aren't that long. Ambi 00:57, 1 December 2005 (UTC)

If we were looking at making stable versions of x articles to print then we could afford to take time time over it. I think we're aiming for different things, here. --Cherry blossom tree 16:38, 1 December 2005 (UTC)

Based on Wikipedia:Trust model, we can have the most ratified version to be publicised and with a disclaimer. So we have a spectrum of trust in the article from Unvandalised version (as you suggest), through to Wikipedia:Standard articles, Wikipedia:Good articles, and Wikipedia:Featured articles, or more. Publicised articles then must have such an explicit disclaimer associated with it. An Unvandalised version may still have inaccurate information so the benefit is not so great. The publicised edition would need at least to achieve something like Standard articles, not high quality, but at least all is factually accurate, which is something you can present to anyone with confidence. -- Zondor 01:53, 1 December 2005 (UTC) -- Zondor 02:07, 1 December 2005 (UTC)

There's no need try to scale this process to cover all of Wikipedia. If we could just publish 10000 reviewed articles on core topics, we'd have something quite good. Fredrik | tc 02:01, 1 December 2005 (UTC)

I strongly agree with Fredrik on this! Over time, we will have quite a respectable subset of Wikipedia articles. In some way, it is a good reward for those who spend a great deal of time and effort on getting articles to FAC! - Ta bu shi da yu 02:47, 2 December 2005 (UTC)

that would be a good strategy. schools don't need pokemon-cruft.-- Zondor 03:18, 1 December 2005 (UTC) Wikipedia:WikiProject Missing encyclopedic articles; Wikipedia:List of articles all languages should have -- Zondor 03:27, 1 December 2005 (UTC)

In any case it would be a good idea to start with a limited list of topics to be reviewed. When there is a process that works, expanding it to the whole of Wikipedia could still make sense in the future (and so schools will be able to enjoy some perfect Pokemon articles). Kusma (talk) 03:35, 1 December 2005 (UTC)

Self-stabilization

Here is a half-baked idea seeking bakers. Large parts of Wikipedia have been very stable. High quality sentences, paragraphs, sections, and occasionally whole articles go untouched for long periods of time. In some cases this is due to lack of interest, but in other cases it is due to stability, i.e. consensus among editors. The idea is that the older text is, the harder it should be to change. The half-baked part of it is that there needs to be a continuously variable stickiness operator that prevents radical changes. The only thing I can think of to handle it might be a tree of edits that propagate upward. People could surf Wikipedia seeing the most stable content by default, or see any level of the tree. Leaf nodes on the tree would work their way up the tree into the article of they stand the test of time. So a leaf sentence about G W Bush's parentage, that might be constantly changed by vandals, would never rise up the tree and would never make up to the topmost level, i.e. never modify the stable article that almost all people would see. On the other hand, a paragraph in an article about some aspect of the French Revolution might be modified by an expert and then reviewed by dozens of editors. As time goes by and it rises in the tree, it would be scrutinized by more and more eyeballs. The longer it passes muster, the greater distance between it and any sudden vandalism which would become a new leaf. The longer it passes muster, also, the closer it would get to being part of the stable article. Part of the baking needed for this idea is a way to make articles responsive to edits, especially current events, or highly verifiable info such as fresh census data. Your thoughts, please. Hu 02:30, 5 December 2005 (UTC)

• I was always under the impression that the law of entropy states that all systems tend toward chaos. I feel that we're a bit like this. I don't believe we are constantly self-stabilising. - Ta bu shi da yu 11:54, 6 December 2005 (UTC)
• You might want to discuss this at Wikipedia talk:Stable versions. -- Jmabel | Talk 07:25, 5 December 2005 (UTC)
  • Erm, you do know that you're referring to this page, right? Ambi 07:34, 5 December 2005 (UTC)
    • Ooops! Running through too many talk pages in too rapid succession pickles the brain. -- Jmabel | Talk 08:29, 5 December 2005 (UTC)

• I'm not sure how the mechanics of this would work, exactly. English prose is not very amenable to having parts missing, or mixing and matching different versions. Especially given how a single word or sentence can completely invert the meaning of the surrounding text. The simple passage of time is not a good metric for reliability, as neglected articles sit around for a long time without being touched. I actually find that articles that have seen a high number of edits tend to be more reliable (except in the case of edit wars). I think you could probably assign a reliability metric to an article (and display that in thermometer or through colorization or something) and then note which parts of it have changed recently (perhaps by colorizing them differently). -- Beland 05:42, 6 December 2005 (UTC)

Giving this a shot

I've unblanked the CUPS page and put it back on the stable versions page. I've done this mainly so we can play around with it to see what people come up with. - Ta bu shi da yu 11:52, 6 December 2005 (UTC)

Since this is part of a general exploration of the idea of stable versions, I think we should have the CUPS example (while having made it clear that this is an example of how the system could work). — Matt Crypto 12:01, 6 December 2005 (UTC)

Indeed. This is why I added that italicised text to the subarticle. - Ta bu shi da yu 12:08, 6 December 2005 (UTC)

Oh, and for the record: I much regret the "get stuffed" edit summary. - Ta bu shi da yu 12:09, 6 December 2005 (UTC)

I'm not entirely sure why Ambi is so insistent that the CUPS example cannot be included. This entire proposal is speculative, and there is no assertion that this is an agreed on process. Therefore, I don't see why the CUPS example is so objectionable, particularly if it's explicitly identified as an example of how this sort of process could be implemented. — Matt Crypto 12:29, 6 December 2005 (UTC)

That's why I'm doing it! I believe that the reason Ambi is against it is because "nothing has been decided". Well, nothing has been decided because nothing has been done! I think we should at least trial ideas. We could talk and talk about this forevermore, but at the end of the day it will mean nothing gets done. The subarticle isn't affecting anyone, so there was never any need to blank it, and even less reason to remove it from this webpage. I've just reverted again, I encourage Ambi to use the talk page to discuss why she is doing this. - Ta bu shi da yu 12:31, 6 December 2005 (UTC)

What are we trialling? All you've put up is a nomination process that nearly exactly mimics FAC and ignores practically all the discussion on this page. I understand that you're very keen to get things underway, but creating a clone of FAC just so we can get some form of this page working expediently really doesn't help anyone. Ambi 13:03, 6 December 2005 (UTC)

Well, I have further subdivided the example subarticle into mandatory and option areas to focus on. And I'm not ignoring discussion, I'm more tinkering with ideas. - Ta bu shi da yu 13:23, 6 December 2005 (UTC)

See below. Adding criteria doesn't much help us - it's still creating a process that's very similar to FAC, and I'm really not sure the "best of Wikipedia" option is the best way of using stable versions on Wikipedia. There's at least five other ways we could go about this, all of which would markedly differ from what we already have and could serve a really useful purpose. Ambi 13:31, 6 December 2005 (UTC)

What the heck is the CUPS page ?? linas 18:21, 7 December 2005 (UTC)

Sections to use in the subarticle

Trying a few things I feel we need for a stable article in the subarticle:

• Structural
• Accuracy
• Objectivity (Neutrality)
• Currency & comprehensiveness
• Readibility
• Verifiability

What do people think? Can anyone think of better sections? Feel free to play with the subarticle. - Ta bu shi da yu 12:19, 6 December 2005 (UTC)

Sigh. And how does this differ from FAC? This is what I mean; you're overriding all the discussion above about how we might actually get a process that does something genuinely useful, and maybe helps us on the way to Wikipedia 1.0, because you're impatient that things have taken a while so far. Ambi 13:03, 6 December 2005 (UTC)

It differs from FAC because FAC articles change over time. This is different because we are discussing a particular revision, not an article that changes over time. - Ta bu shi da yu 13:09, 6 December 2005 (UTC)

But it's exactly the same process. By doing this, we may as well just make a protected version of every featured article when it is first featured. It seems to me to be the most useless way that we could possibly implement stable versions of articles, when with a bit more planning, we could do something different which could really move things forward for the project as a whole, and potentially kickstart Wikimedia 1.0. Ambi 13:15, 6 December 2005 (UTC)

What do you have planned, out of interest? Anyway, it's a good issue, and not one that has really been expressed well in this talk page. My idea, now, is to subdivide the subarticle into mandatory criteria and optional criteria. We basically use the most stable revision as the revision that is most accurate and readable. Neutrality is obviously non-negotiable. Comprehensiveness is not necessary, though desirable. - Ta bu shi da yu 13:17, 6 December 2005 (UTC)

You're missing the point. You're still conceptualising this process entirely in the grain of FAC; a bunch of criteria, presumably similar to there, which then lead to the article being protected in another space; things which could still be achieved by just protecting copies of newly promoted FAs. Furthermore, proceeding with this as it currently stands, however, without answering the questions above about scalability and what such a process would hope to achieve seems to me to be not the most helpful way of going about things; it's putting having some process up and running ASAP over actually making sure that that process does something useful. Ambi 13:31, 6 December 2005 (UTC)

Ambi -- I presume that you oppose the idea of Stable Articles being exclusively "Super Featured Articles"? I'm not so keen on that, either. I would prefer to see stable versions being revisions of articles that, while correct and decently written, might still be lacking in areas (e.g. completeness, or illustration). Still, I think Ta bu shi da yu's suggestion is interesting and worthwhile, because any process to establish a stable version might well involve things like "accuracy" and "readibility" etc, even if it's not meant to be "the best of Wikipedia" (which is what I think FAs should be). — Matt Crypto 13:20, 6 December 2005 (UTC)

You're basically right in that I don't see much point in this being yet another process for determining good articles; the problem with TBSDY's proposal so far is that it bears remarkable similarity to FAC. The main problem thusfar is that people haven't agreed on precisely what this page is supposed to achieve. Is this to be another "best of Wikipedia" page? Are we going to try and get as much of Wikipedia as possible covered in these stable versions? Are we going to try and use this as a stepping stone to Wikipedia 1.0? Are these stable versions going to just be protected from vandalism, or are we going to make sure that they're accurate as well? All of those goals would require different methods of going about it, which is why I see trying to proceed with one specific one straight away as not being very helpful. Ambi 13:31, 6 December 2005 (UTC)

Well, let's hear it. What are your ideas on getting a stable version into play? No offense intended here, but I haven't actually heard an original idea as yet. I really am asking this to find a better way! If you have one, I'm all ears. - Ta bu shi da yu 13:34, 6 December 2005 (UTC)

The way that I would personally proceed would probably be to try to use this to advance Wikipedia 1.0. Briefer versions of articles (suitable for print) could be added and approved. As they would be much shorter than most good Wikipedia articles, they would be much easier to verify (solving a major problem), and much easier to finetune from current content relatively quickly (dealing with issues of scalability). They would have the advantage of contributing to a clear use down the line (Wikipedia 1.0). There are plenty of other options, though, and this may well not be the best. I'm very much open to suggestions. Ambi 13:38, 6 December 2005 (UTC)

An admirable idea, but not one presented by this proposal. This proposal is to get the whole article, in all its gory detail up to speed so that it can be relatively stable and people can have some confidence that it hasn't been vandalised or hold any really bad inaccuracies. - Ta bu shi da yu 13:41, 6 December 2005 (UTC)

This may be what you've had in mind, but there's been plenty of other ideas floated by other people here that may be better ways of doing this. Indeed, Zondor, who created this page in the first place, has explicitly suggested that this process be linked to Wikipedia 1.0.

Is it worth putting all this effort into making sure articles aren't vandalised, when (with the exception of articles like George W. Bush), this isn't a major issue most of the time? While the addition of an accuracy requirement may make it more useful, how would you make sure that an article doesn't contain any really bad inaccuracies? Unless a much stronger system of fact checking was added, I suspect that such a system would be a lot of effort for comparatively little gain - it would be no more of a reliable source than the regular Wikipedia, and indeed would just be more out of date. Ambi 13:46, 6 December 2005 (UTC)

I have asked in Wikipedia talk:Featured article candidates to slightly change their procedure to make their way into this proposal. -- Zondor 14:32, 6 December 2005 (UTC)

My hope for "Stable versions" is that we'll have a simple, lightweight process. The actual criteria as to whether a version is "stable" is left up to those who are making the actual decision. I beleive that trusting the good intentions of the folks voting to declare a stable version is sufficient. Thus objections about neutralty, accuracy, etc. should be sufficient to block the declaration of a stable version. However, if no objections are raised, then an aricle may be declared stable (even if it is known to be lacking in various ways). I'm proposing something subtle here: that objections about POV are enough to stop declaration of stability. This is different from the laborius, intensive process required to vet an article to be truly, absolutely free from POV. linas 18:37, 7 December 2005 (UTC)