Wikipedia:Stabilizing featured articles

There is currently a discussion among members of the Wikipedia community as to whether Wikipedia should begin to "stabilize" articles after a certain point in their development &mdash; that is, static versions of articles which have gone through some sort of review process, and are periodically replaced with versions developed by the standard, collaborative methods upon which Wikipedia has always relied. Such stabilization gives Wikipedia the ability to point definitively at articles and say that their quality has been certified, the articles can be trusted, and even vandals cannot diminish this status.

Such proposals meet with the criticism that it will slow article development, will further divide administrators who will be able to edit stabilized articles from other editors who won't, and will make the entire contribution process far more complicated. The greatest concern, however, is that stabilizing articles in such a manner that changes need to be requested before they are seen will limit the freedom that draws many editors here in the first place, the lure of a massive encyclopedia that anyone can edit.

This proposal aims to achieve the benefits of stabilization while addressing the concerns of the community, all while using the software we already have in place.

Why stabilize?
In general, all stabilization proposals seek to increase the trustworthiness of Wikipedia articles. Because of the dynamic nature of the project, it is asserted that only static versions that have undergone peer review can be fully trusted. Most proposals also recognize the importance of keeping some degree of the dynamic, Wiki process employed to drive article improvement forward and, more generally, to maintain the core, founding spirit of the project. Because there already is a peer review process at Wikipedia to create Featured Articles, stabilization can dovetail with that process. Good Articles, which also undergo peer review, could be considered for stabilization as well.

Using the history function to certify the quality of an article
The truth is, we already have stabilized versions of every single article on Wikipedia: the article history. Every bit of brilliant prose written in Wikipedia's history can be accessed as easily as current versions of articles. We use this strength to our advantage.

An example: Linus Pauling
I have set up an example at User:JDoorjam/Linus Pauling. The Pauling article was given Featured Article status on July 12, 2006. After thorough peer review, editors of Wikipedia agreed that the Pauling article was one of the best in the encyclopedia, and granted it featured status. But of course, the encyclopedia is a dynamic one, and continued to evolve. Under this proposal, a banner at the top of the page will explain that the article was deemed good enough to be a Featured Article, and if one wishes to see the Certified version, one is only a click away. Meanwhile, if one is interested in plunging in and making the article even better, one can do so as always. And another link on the banner shows what has changed since the article was certified Featured.

Updating the "Featured" version
After some period of time, editors of a Featured article may decide they've made significant improvements and would like to update the "Featured" version. (This happens currently as well, typically as either part of article improvement drives, or if an article has lost featured status and editors wish the article to regain it.) All editors must do is submit it for re-review. This could be done in the same way it's done now. Recertifying would presumably be a much simpler process than the first certification, provided the article has maintained its quality. The peer review process could be as simple as comparing the two versions by clicking the "Compare versions" list, and citing any areas of concern. After a (perhaps truncated) review process, the version deemed "certified" is updated.

Benefits
There are a number of benefits that this method yields.


 * 1) It's simple. After an article is certified "Featured", a banner is put on the top of the article, just like any other banner on a Wikipedia article.  Editing continues as normal; no one has to learn a new system.
 * 2) It uses the software we have now. Changes may be implemented down the road, but if a system such as this one is in place, hopefully any changes made will use this or a very similar format, easing transition and not losing editors in the shuffle.
 * 3) The dynamic version is the default version. The most major concern about stabilizing, that the Wikipedia "anyone can edit" spirit will suffer, is addressed here.  Anyone can still edit as always.  No one need wait for a sysop to approve their change: it will still be seen right away.  And, as always, there's still the goal of article improvement to strive for.  The difference is that now, when a group of editors reaches the point where an article is good enough to be Featured, that record is kept, like a high score against which editors can compare future edits.
 * 4) The edit history is preserved. There is no forking in the article, no article or talk page splitting, and every edit is preserved and linearly recorded, as it is now.  Even though there is a "Certified" version and a "Dynamic" version, there is still only ever one article thread.

Technical issues
With the idea behind the proposal stated above, there are several implementation suggestions offered here.

Keeping track of the banners, and protecting them from vandalism
A concern has been stated that, while vandals could not alter the "Certified" version, they could tamper with the banner at the top of the page. There is, after all, no way to lock part of a page. Thus, rather than being code directly on the page, for reasons of both security and ASCII aesthetics, locked templates should be transcluded from template space into the top of the article. For instance, the template for Linus Pauling would reside at Featured articles/Linus Pauling. While vandals could potentially remove the banner from the top, such vandalism would be noticed immediately and reverted, the same as it would if a featured article banner were removed from the talk page of an FA. As an extra layer of security, if warranted, it would be a trivial task for a bot to monitor the FA templates to make sure each was transcluded once and only once.