Knowledge Base changes: Difference between revisions

From MozillaZine Knowledge Base
Jump to navigationJump to search
(→‎Dispute resolution process: reply to proposal by Tanstaafl)
Line 89: Line 89:
The above outline basically expresses my point of view, that a dispute resolution process should start with the individuals talking to each other and backing off from further editing, then advance to informal mediation, discussion with third parties, and request for mediation.  Arbitration, in which an individual or committee ''will consider the case and issue a decision, instead of merely assisting the parties in reaching an agreement'', should only be taken as a last resort.  More references: [http://en.wikipedia.org/wiki/Wikipedia:Edit_war Wikipedia:Edit war]; [http://en.wikipedia.org/wiki/Wikipedia:Why_Wikipedia_is_not_so_great Wikipedia:Why Wikipedia is not so great] with  sections on [http://en.wikipedia.org/wiki/Wikipedia:Why_Wikipedia_is_not_so_great#Quality_of_existing_entries Quality of existing entries] and [http://en.wikipedia.org/wiki/Wikipedia:Why_Wikipedia_is_not_so_great#Collaboration_practices_and_internal_social_issues Collaboration practices and internal social issues]. [[User:Alice Wyman|Alice Wyman]] 13:00, 8 December 2006 (UTC)
The above outline basically expresses my point of view, that a dispute resolution process should start with the individuals talking to each other and backing off from further editing, then advance to informal mediation, discussion with third parties, and request for mediation.  Arbitration, in which an individual or committee ''will consider the case and issue a decision, instead of merely assisting the parties in reaching an agreement'', should only be taken as a last resort.  More references: [http://en.wikipedia.org/wiki/Wikipedia:Edit_war Wikipedia:Edit war]; [http://en.wikipedia.org/wiki/Wikipedia:Why_Wikipedia_is_not_so_great Wikipedia:Why Wikipedia is not so great] with  sections on [http://en.wikipedia.org/wiki/Wikipedia:Why_Wikipedia_is_not_so_great#Quality_of_existing_entries Quality of existing entries] and [http://en.wikipedia.org/wiki/Wikipedia:Why_Wikipedia_is_not_so_great#Collaboration_practices_and_internal_social_issues Collaboration practices and internal social issues]. [[User:Alice Wyman|Alice Wyman]] 13:00, 8 December 2006 (UTC)
:I believe 1-4 are essentially what we allow for already. Of course, we don't conduct surveys, but discussion on talk pages would allow for such a thing to happen. 5 is what I'm proposing, but I worry that if we put in arbitration as an option but don't have any standards for the arbitrator to go by, it could seem like the arbitrator is solely speaking his or her own view, or even worse, showing favoritism.--[[User:Np|Np]] 20:54, 8 December 2006 (UTC)
:I believe 1-4 are essentially what we allow for already. Of course, we don't conduct surveys, but discussion on talk pages would allow for such a thing to happen. 5 is what I'm proposing, but I worry that if we put in arbitration as an option but don't have any standards for the arbitrator to go by, it could seem like the arbitrator is solely speaking his or her own view, or even worse, showing favoritism.--[[User:Np|Np]] 20:54, 8 December 2006 (UTC)
::I know you're proposing arbitration but it's a heavy-handed method no matter how unbiased the arbitrator(s) may be, which is why it is left as a "last resort".  Dispute resolution is a delicate area and it takes time in any open society or organization.  I'm proposing that a [http://en.wikipedia.org/wiki/Wikipedia:Request_for_comments requesting comments from other editors] and/or a formal [http://en.wikipedia.org/wiki/Wikipedia:Dispute_resolution#Mediation mediation process] take place first, as a required step before arbitration, and that both mediation and arbitration are steps that the parties must request.  If you just want to stop an edit war, can't you set up a "maximum revert" quota (Wikipedia uses three reverts, in whole or in part,  to a page within 24 hours) and institute a block for further edits against any user where the quota is exceeded?  See http://en.wikipedia.org/wiki/Wikipedia:Three-revert_rule#Enforcement  [[User:Alice Wyman|Alice Wyman]] 21:51, 8 December 2006 (UTC)
::I know you're proposing arbitration but it's a heavy-handed method no matter how unbiased the arbitrator(s) may be, which is why it is left as a "last resort".  Dispute resolution is a delicate area and it takes time in any open society or organization.  I'm proposing that [http://en.wikipedia.org/wiki/Wikipedia:Request_for_comments requesting comments from other editors] and/or a formal [http://en.wikipedia.org/wiki/Wikipedia:Dispute_resolution#Mediation mediation process] take place first, as a required step before arbitration, and that both mediation and arbitration are steps that the parties must request.  If you just want to stop an edit war, can't you set up a "maximum revert" quota (Wikipedia uses three reverts, in whole or in part,  to a page within 24 hours) and institute a block for further edits against any user where the quota is exceeded?  See http://en.wikipedia.org/wiki/Wikipedia:Three-revert_rule#Enforcement  [[User:Alice Wyman|Alice Wyman]] 21:51, 8 December 2006 (UTC)


:::Edited to restore Alice's comment, which got accidentally removed when editing jobs crossed in the mail.  [[AnotherGuest.]]  8 Dec 06
:::Edited to restore Alice's comment, which got accidentally removed when editing jobs crossed in the mail.  [[AnotherGuest.]]  8 Dec 06
Line 101: Line 101:


::::This is expensive, slow, and might not pick the best choice but one way or another it would force a resolution without any suspicion of favoritism by the other editors. Whatever we decide should be sanity checked against a couple of different cases. It might not always be multiple authors reverting each others edits. It might be one-sided with one author changing the text and everybody else just reverting his changes because they thought they were inappropriate. This occurred in the anti-virus article where somebody essentially kidnapped it to argue at great length against running anti-virus software. [[User:Tanstaafl|Tanstaafl]] 08:12, 9 December 2006 (UTC)
::::This is expensive, slow, and might not pick the best choice but one way or another it would force a resolution without any suspicion of favoritism by the other editors. Whatever we decide should be sanity checked against a couple of different cases. It might not always be multiple authors reverting each others edits. It might be one-sided with one author changing the text and everybody else just reverting his changes because they thought they were inappropriate. This occurred in the anti-virus article where somebody essentially kidnapped it to argue at great length against running anti-virus software. [[User:Tanstaafl|Tanstaafl]] 08:12, 9 December 2006 (UTC)
:::::Regarding your first point, on asking a Sysop to limit reversions or lock down an article in response to an edit war:  We'll need to set a policy on limiting number of reversions, probably by voting on it,  since it's not something that a Sysop should decide on his own, in my opinion.  Since we're such a small group, the limit would probably need to be much stricter than 3 reverts per 24 hours as in my Wikipedia example, which includes a HUGE pool of editors.  In our tiny little pond,  maybe three reverts per week to the same article would work better, but it needs discussion, as does the penalty and duration of the sanction (blocking the editor, protecting the page or whatever).
::::: I agree with your proposal in general and I especially like the idea of asking the disputing authors to create different versions of the same article and to gather opinions on the forums.    I've proposed new articles in the past by placing the article on my [[:User:Alice_Wyman/Proposed_article]] page, and then linking to the proposed article on a discussion page.  I've also PM'ed people on the forums with a link to the article to get more feedback.  I also like the idea of potentially moving mediation a or arbitration off the KB and onto the forums, especially in cases where the dispute is over technical content, since the forums have  many more potential contributors.  To add to your proposal:
:::::* I would specify how the formal dispute resolution process is initiated.  I think that at least one of the editors participating in the disputed edits should request mediation before the Sysop is involved, unless the conduct becomes so egregious that intervention has been requested by two or more observers to the dispute.
:::::* Once formal mediation is requested, all parties in the dispute should agree to take part in the dispute resolution process before it can proceed.  For example,  No one should be "forced" to produce a proposed version of the disputed article.  If two or more proposed articles are produced, it will go to dispute resolution.  If only one party produces a proposed version of the article, that version will be accepted, and it will be assumed that the party that did not produce a proposed article has conceded.  In that case, the conceding editor should  be banned from reverting the article to the disputed version;  however, he should not be prevented from future "constructive" edits.
:::::*  Once the dispute resolution process begins, the parties should agree to stop all editing of the disputed portions of the article.    until the dispute is resolved. Another option would be to "lock down" or place the article off limits to all edits, even from parties not involved in the dispute, as a means to encourage quick resolution through social pressure (this would be an extreme measure which should probably be decided by consensus or vote, e.g. an "articles nominated for lockdown" process).  The discussion should then be moved off of the article Talk page to a separate "dispute resolution" page or to a forum thread if one or more chosen mediators don't have a KB account.
:::::* I don't think that the Sysop should limit the pool of KB editors or forum members who would comment on the proposed article versions, since this can also be considered to be part of the mediation process.  Gathering comments and ideas from registered forum members may  encourage the disputing editors to come to a resolution on their own. 
:::::* If the parties still cannot resolve the dispute, only then should the Sysops as a group select a "arbitration panel" (optionally, the panel may already be in place, as a standing "Arbitration Committee") or another option would be for the disputing parties to choose their own panel to decide the matter.
:::::I know that the above is going into way too much detail and it goes without saying that this is a slow and "expensive" process, in that it takes up a lot of user resources.  The most important part of the process is that it should be as free as humanly possible from  "political" considerations such as protecting bruised egos, coddling new editors or backing up prolific editors to keep them contributing,  currying favor with a forum Moderator or popular forum member (one of the downsides of going to the forums) and the like.  Another important consideration is to keep the Sysops out of the actual decision-making process as far as possible, to avoid the criticism that they are abusing their power.  [[User:Alice Wyman|Alice Wyman]] 16:20, 9 December 2006 (UTC)


==KB user feedback==
==KB user feedback==

Revision as of 16:20, 9 December 2006

This page has been created for several reasons.

  • It would be nice to have a place where new editors can introduce themselves and meet existing editors.
  • It would be good to allow new editors to safely propose content changes (minor or major) prior to implementing them.
  • It would be good to have a central location to discuss the style, content and organisation of this Knowledge Base. (Some of the ideas in Talk:Knowledge Base can be migrated here, leaving that page solely for discussion of the front page article itself.
  • This page was an attempt to address incidents that have occured on the KB where some groups of editors have been unaware of major changes being made by other groups of editors.

This page is the primary place to announce new suggestions. Whenever possible, issues should be discussed in a more appropriate place, such as the discussion page of the article or category that the suggestion affects.

Once suggestions are resolved, they are moved to Knowledge Base changes/Archive.

Welcome to new editors

Hello! Great to have you here. Please add a comment here :-)

  • Hello Knowledge Base! FJ reporting to duty! *bows at all* FatJohn 11:30, 3 January 2006 (UTC)
  • I guess I'm new, I am "name already taken" from the forums, hello to everyone! --Lethargy 21:34, 10 May 2006 (UTC)

Ownership and dispute resolution

Prompted by the comments in this forum thread, I believe we need to some sort of dispute resolution mechanism and general guidelines in the KB. Otherwise, disagreements can easily escalate into edit wars and contributors leaving. The obvious solution is to have the Sysops handle these things, but I don't know if there are enough active Sysops to present a decision ("The Sysops said so, so I'll do it" vs. "Np is crazy, screw him"). Thoughts?--Np 21:43, 6 December 2006 (UTC)

I also noticed that thread and it wasn't a nice read. I've been overrun a few times myself. Usually with an accompanying explanation though. Maybe you need more sysops then. If a policy is made, it should be promptly introduced to new editors to make sure they're aware of it. PS. I wish all KB editors should "watch" this page. Could it be automatically inserted into new editors watchlist? --FatJohn 01:28, 7 December 2006 (UTC)
As of right now, the Sysops don't have the ability to change people's watchlists. Administrators might.--Np 03:23, 7 December 2006 (UTC)
First, what do you mean by "ownership"? Should articles be "owned" by different editors or Sysops who have the final say on edits, for example, give Vectorspace all the plugin articles, Tanstaafl all the Thunderbird articles and you and Unarmed take the preferences articles? Or did I misunderstand?
No, what I mean is that someone or a group of people is "in charge" of the Knowledge Base, to create and modify policy and resolve disputes between members. There would not be anyone in charge of particular articles; members would be free to edit as they currently do (of course, with the exception of the dispute resolution process and any new policy created). --Np 03:23, 7 December 2006 (UTC)
Second, I think that the current system of resolving editing disputes by posting to the article's Talk page should work in 99% or the cases. Hopefully, other editors would add to the discussion but if not, or if the issue still isn't resolved after a reasonable period, I suppose a Request for mediation page might work, something similar to the Pages voted for deletion discussion? Or, maybe one of the involved parties could simply post a "Mediation request" here, on this page? Once such a request is made, further editing of the page by the disputing parties should cease until the editing conflict is resolved. Such a mediation request could also be looked at as a request for "quality review" of the article, since after all, what we really want is for the article to be well written, understandable by the intended reader and technically accurate. I mentioned in the forum thread that some sort of "KB quality control" process is needed, besides just hoping that errors are caught and fixed as people review the recent changes or happen to read an article and spot a problem. Alice Wyman 01:59, 7 December 2006 (UTC)
Yes, the talk pages are usually sufficient in resolving disputes, though I suspect that many of disputes that arise are dropped by one of the parties to avoid the edit war. With a dispute resolution system, people wouldn't have to fear alienating users or entering an endless argument by asserting their views.
If you don't want people to risk alienating other editors then maybe a Request for quality review of the article (made by either one of the disputing parties OR by anyone else observing the conflict) would be the better route, but first a quality review process should be in place, hopefully similar to the review process taken before deleting an article, where everyone has the opportunity to contribute. I guess what I'm trying to get at is that a mediated compromise is preferable to an arbitrary Sysop decision (in the sense of mediation as opposed to arbitration). Of course it's easier for a Sysop to dictate what the resolution should be but I think that an open process where all editors are free to participate is better in the long term. Alice Wyman 12:00, 7 December 2006 (UTC)
The ability to change policy would help with quality control - we could dictate by policy in general terms what an article should or shouldn't be like.--Np 03:23, 7 December 2006 (UTC)
There are already general guidelines given in Rules#Editing_courtesy about documenting controversial changes (the linked Talk page even mentions using the Talk pages to suggest "Possible solutions to edit wars" in the "How to use Talk" section) so maybe that's where dispute resolution should be mentioned, but nothing specifically about quality guidelines, I guess because it's a difficult subject. Alice Wyman 12:00, 7 December 2006 (UTC)
We already have the ability to add a tag stating that the article doesn't meet wikipedia standards. The few times I've seen it used it didn't cause a dispute because it was pretty obvious that it deserved it. Perhaps the real issue is what types of disputes shouldn't be left to the talk pages. I remember the great difficulty several editors had dealing with one person who was basically kidnapping the anti-virus article to evangelize not using anti-virus software. That wasn't just two editors having a dispute, it was a case of whether one of the editors was acting inappropriately. Tanstaafl 15:47, 7 December 2006 (UTC)
The issue isn't messy pages. We have plenty of standards for grammar and formatting, but none about content. For example, should support articles present easy information to users to get them going again, or should they explain things in detail? In what order should solutions be posted - easiest first or most likely first? Should solutions be referenced so we know editors aren't just making stuff up? This is the type of policy we don't have that I think we desperately need. With an official policy on these things, users would be more comfortable changing "other people's" articles.
Regarding the dispute you mention, this is exactly where dispute resolution would come in. Right now, the only way to "win" a dispute is to convince the other editor or to be more persistent in implementing your version. With a dispute resolution process, the latter is replaced with an official process that takes less time, is less alienating and frustrating, and is more likely to yield the "correct" solution.--Np 18:18, 7 December 2006 (UTC)
"For example, should support articles present easy information to users to get them going again, or should they explain things in detail?"
Both, but sometimes often there are no easy answers because there are too many variables or unknowns. Sometimes the best that can be done is to explain how things work and suggest a check list or some guidelines for troubleshooting.
"Should solutions be referenced so we know editors aren't just making stuff up?"
Yes and no. What would you reference for the Standard Diagnostic? That's more of a messy consensus than a referenceable procedure. Although it may not be efficacious to reference a procedure, information should usually be referenced unless it's common knowledge and beyond dispute. --AnotherGuest 7 Dec 06
I wasn't intending on actually forming the policy in this discussion. I just want to figure out whether a defined policy would be useful, and who would create it.--Np 20:47, 7 December 2006 (UTC)
I am uncomfortable with the direction you're advocating. I don't mind a honest disagreement with other editors but I've repeatably had problems with what I consider grammar/style guide nazis who don't use any judgement, do their damage, and move on without contributing anything to the article. On the face of it there is nothing wrong with some of the standards you're proposing but based on past history it will just make things worse.
You mention With an official policy on these things, users would be more comfortable changing "other people's" articles. I'm not convinced thats a big problem. What I have noticed is we're not getting enough new articles about users problems (reference articles about a specific preference don't count) and too many cases where a ordinary user writes one article and then decides they've had it with the knowledge base.
I consider our biggest problem to be lack of contributers, not how professional the editors are, or how willing they are to edit somebody elses work. There appears to be a core of about six or seven individuals that do the majority of the work. I find a wiki style of collaboration to be difficult at times since you don't own any articles you write. I've found I usually have no problem when just one editor edits whatever I've written, regardless of how extensive the changes are or the differences in styles (as long as its clear they're not making changes just for the sake of standards). It gets more difficult the more editors are involved because there is no longer one consistent writing style. That can reduce the quality more than any deficiencies in my writing skills. It also creates complications because if you get too aggressive on rewriting an article the other contributers may feel they're being edited out. I can deal with that. However, I have a problem when you combine that with "editors" that don't contribute anything to articles for that application but want to dictate in detail many decisions.
Yes, lack of contributors is a major problem, but I'm not trying to address that with this discussion.--Np 20:54, 8 December 2006 (UTC)
I'm used to collaborating with some of the more prolific helpers in the forums. I'm struggling to find a way to encourage users who don't want to write/edit an article to contribute/pool information that I'll use to write an article. For example, I've been trying to get users that I've been helping with problems with the U3 version of Thunderbird to post background information such as the USB drive layout. However, I basically just want to be left alone by other editors, unless they're going to actually contribute to an article. I've noticed that the "standards" are in practice dictated by a few editors who hold similar views, not by consensus of all of the editors. I don't consider somebody mindlessly editing an article to make it conform to some standard that I didn't agree with in the first place as adding any value. Tanstaafl 04:32, 8 December 2006 (UTC)
The standards would not be arbitrary nitpicky things, much less nitpicky than the current in house style. They would instead be major issues that affect how useful an article is to a user. An example of a standard for an "issue" page could be "Start with a short description of the article. What are the common symptoms of the problem? What other problems are related to this problem, but are covered elsewhere? Then, as plainly as possible, describe the steps to solve the problem. Then, provide technical information for the more advanced users".
Furthermore, this would not be "Let's make Tanstaafl write articles how Np wants". This would be "Let's get all the editors get together, corroboratively decide the standards that would help users, then work on making articles fit those standards."--Np 20:54, 8 December 2006 (UTC)
You haven't convinced me there is a real problem while I am well aware of potential downsides. I prefer the current solution which is to let other editors improve articles as needed and lead by example. Tanstaafl 09:05, 9 December 2006 (UTC)

Quality control process

I wouldn't want to set any specific standards for what to include in articles such as should support articles present easy information to users to get them going again, or should they explain things in detail, or whether to place the easiest versus the more common solutions first since you can't really generalize. Like I said before, what we really want is for the article to be well written, understandable by the intended reader and technically accurate. Including references in articles not only helps the reader but helps anyone reviewing the article verify accuracy. When I suggested a quality control process I meant that a review process should be set up so that articles get updated or corrected systematically. For example, the About:config entries article needs updating for Firefox 2 menu option changes, since right now it is not technically accurate in some instances as menu options are changed, as are prefernce default settings. Correct me if I'm wrong but right now, articles are only updated or reviewed if someone comes across it and notices something amiss, or if it comes up in the recent changes list. That's why I mentioned a feedback process for non-editors, so that errors or problems get noticed. On way that articles get reviewed as a part of a group effort right now is if someone nominates it for deletion. I would at least like to see a system set up so that articles can be nominated for quality review. Nominating an article for quality review or requesting comments from other editors about article content could also be part of a dispute resolution process in the case of conflicting opinions about what information should be included or removed from an article or how the article content should be organized. Maybe some editors could look at the Lost bookmarks article and it's Talk page as a case in point? Alice Wyman 02:23, 9 December 2006 (UTC)

I periodically browse most of the Thunderbird articles looking for ones that seem out of date. That doesn't mean I'll do something about every one I find (I know the bouncing messages article needs updating for 1.5 for example, but I keep dropping it to the bottom of my list of things to do) but I do make the effort. I assumed other people did the same thing for areas that they concentrated on.
You mention two different reasons - a poorly written article and its out if date. Your example seems more like a poster child for dispute resolution (between three editors) than a quality review process. Yes, there's some overlap but I think we're better off treating dispute resolution as a separate problem.
I suggest we start small and maintain a list of articles that are out of date, that includes a one line description of the problem. Thats less likely to cause disputes or hurt feelings and makes it easier to bring to somebodies attention since they don't have to scan the articles looking for a update tag. If the process works well then we could expand it (in several iterations) to deal with whether its well written or understandable (which is a much trickier issue to deal with since that may involve dispute resolution, standards, how good user feedback is, and political minefields such as the profile folder article) later on. I'm basically arguing for a iterative process (to discover and/or adapt whatever meets our needs), rather than mimicking whatever the Wikipedia does since we have only a few dozen editors while they have over 2.5 million.
The Thunderbird and Firefox support forums have a FAQ sticky thread (I don't know why SeaMonkey doesn't). I (or somebody else) could add a short blurb that links to that page, a similar blurb to the top of http://kb.mozillazine.org/Knowledge_Base and a thread in the general forums about it (mainly to let the more prolific posters know about it and ask them to mention it as needed). Tanstaafl 06:03, 9 December 2006 (UTC)

Dispute resolution process

I agree with Tanstaafl that our biggest problem is lack of contributors, since a wide base of editors would quickly resolve disputes, either through timely correction of bad writing or content errors or by other editors adding input to the discussion. I did some research using http://en.wikipedia.org/wiki/Wikipedia:About as my starting point, and found a section on Handling disputes and abuse that we can use to form our own dispute resolution process. Here is how Wikipedia handles editing disputes:

Unresolved disputes between editors, whether based upon behavior, editorial approach or validity of content, can be addressed through the talk page of an article, through requesting comments from other editors or through Wikipedia's comprehensive dispute resolution process.

The "requesting comments" link says, Requests for comment (RfC) are the open part of the dispute resolution process, by which editors can seek broad input regarding disputes over article content, user conduct, and Wikipedia policy and guidelines. Here is an outline from the dispute resolution page, which is tagged as "an official policy on the English Wikipedia."

  • 1 Avoidance
  • 2 First step: Talk to the other parties involved
  • 3 Second step: Disengage for a while
  • 4 Further dispute resolution
    • 4.1 Informal mediation
    • 4.2 Discuss with third parties
    • 4.3 Conduct a survey
    • 4.4 Mediation
  • 5 Last resort: Arbitration
  • 6 Requesting an Advocate (at any time)

The above outline basically expresses my point of view, that a dispute resolution process should start with the individuals talking to each other and backing off from further editing, then advance to informal mediation, discussion with third parties, and request for mediation. Arbitration, in which an individual or committee will consider the case and issue a decision, instead of merely assisting the parties in reaching an agreement, should only be taken as a last resort. More references: Wikipedia:Edit war; Wikipedia:Why Wikipedia is not so great with sections on Quality of existing entries and Collaboration practices and internal social issues. Alice Wyman 13:00, 8 December 2006 (UTC)

I believe 1-4 are essentially what we allow for already. Of course, we don't conduct surveys, but discussion on talk pages would allow for such a thing to happen. 5 is what I'm proposing, but I worry that if we put in arbitration as an option but don't have any standards for the arbitrator to go by, it could seem like the arbitrator is solely speaking his or her own view, or even worse, showing favoritism.--Np 20:54, 8 December 2006 (UTC)
I know you're proposing arbitration but it's a heavy-handed method no matter how unbiased the arbitrator(s) may be, which is why it is left as a "last resort". Dispute resolution is a delicate area and it takes time in any open society or organization. I'm proposing that requesting comments from other editors and/or a formal mediation process take place first, as a required step before arbitration, and that both mediation and arbitration are steps that the parties must request. If you just want to stop an edit war, can't you set up a "maximum revert" quota (Wikipedia uses three reverts, in whole or in part, to a page within 24 hours) and institute a block for further edits against any user where the quota is exceeded? See http://en.wikipedia.org/wiki/Wikipedia:Three-revert_rule#Enforcement Alice Wyman 21:51, 8 December 2006 (UTC)
Edited to restore Alice's comment, which got accidentally removed when editing jobs crossed in the mail. AnotherGuest. 8 Dec 06
I agree that arbitration is heavy-handed and a last resort. I'm just unsure whether anyone would be interested in commenting on a dispute who wouldn't already comment on the talk page.--Np 03:44, 9 December 2006 (UTC)
Temporarily stopping the edit war only buys more time. That should be something that anybody could ask a SysOp to do (though they might say no, apply a limit on the number of reversions, temporarily block the feuding authors from editing the article, or protect the page). Perhaps we could use the following process once it deteriorates to a edit war.
1. A SysOp tries to arrange formal mediation by a neutral person that the authors agree on. The mediator might solicit comments from editors, helpers and/or users, conduct a survey etc.. If the SysOp can't arrange mediation they will wait at least two weeks for things to cool down and see if the problem resolves itself before moving to the next step.
2. If that fails (for whatever reason) the SysOp will ask the different authors to produce independent versions of the same article, create a thread to discuss it in the appropriate general forum (so that its limited to only registered users), and ask the most well known helpers (that hang out in that forum) to critic them. The SysOp should provide at least a week for the authors to create their own version, and at least a week for any registered user to critic the articles (and apply social pressure). If a author refuses to provide their own version of the article than they will be permanently blocked from editing the article.
3. If that fails to resolve the dispute the SysOp will create another thread in the same forum and have a formal poll for at least a week to pick the winner. Once the winner is chosen the thread will be locked. If the SysOp isn't a forum moderator they will ask a moderator to create the poll and lock the thread afterwards. It is the SysOps (not the moderators) decision when to declare the winner. If a author refuses to accept the decision than the SysOp should ban them for the appropriate amount of time. That might be permanently.
This is expensive, slow, and might not pick the best choice but one way or another it would force a resolution without any suspicion of favoritism by the other editors. Whatever we decide should be sanity checked against a couple of different cases. It might not always be multiple authors reverting each others edits. It might be one-sided with one author changing the text and everybody else just reverting his changes because they thought they were inappropriate. This occurred in the anti-virus article where somebody essentially kidnapped it to argue at great length against running anti-virus software. Tanstaafl 08:12, 9 December 2006 (UTC)
Regarding your first point, on asking a Sysop to limit reversions or lock down an article in response to an edit war: We'll need to set a policy on limiting number of reversions, probably by voting on it, since it's not something that a Sysop should decide on his own, in my opinion. Since we're such a small group, the limit would probably need to be much stricter than 3 reverts per 24 hours as in my Wikipedia example, which includes a HUGE pool of editors. In our tiny little pond, maybe three reverts per week to the same article would work better, but it needs discussion, as does the penalty and duration of the sanction (blocking the editor, protecting the page or whatever).
I agree with your proposal in general and I especially like the idea of asking the disputing authors to create different versions of the same article and to gather opinions on the forums. I've proposed new articles in the past by placing the article on my User:Alice_Wyman/Proposed_article page, and then linking to the proposed article on a discussion page. I've also PM'ed people on the forums with a link to the article to get more feedback. I also like the idea of potentially moving mediation a or arbitration off the KB and onto the forums, especially in cases where the dispute is over technical content, since the forums have many more potential contributors. To add to your proposal:
  • I would specify how the formal dispute resolution process is initiated. I think that at least one of the editors participating in the disputed edits should request mediation before the Sysop is involved, unless the conduct becomes so egregious that intervention has been requested by two or more observers to the dispute.
  • Once formal mediation is requested, all parties in the dispute should agree to take part in the dispute resolution process before it can proceed. For example, No one should be "forced" to produce a proposed version of the disputed article. If two or more proposed articles are produced, it will go to dispute resolution. If only one party produces a proposed version of the article, that version will be accepted, and it will be assumed that the party that did not produce a proposed article has conceded. In that case, the conceding editor should be banned from reverting the article to the disputed version; however, he should not be prevented from future "constructive" edits.
  • Once the dispute resolution process begins, the parties should agree to stop all editing of the disputed portions of the article. until the dispute is resolved. Another option would be to "lock down" or place the article off limits to all edits, even from parties not involved in the dispute, as a means to encourage quick resolution through social pressure (this would be an extreme measure which should probably be decided by consensus or vote, e.g. an "articles nominated for lockdown" process). The discussion should then be moved off of the article Talk page to a separate "dispute resolution" page or to a forum thread if one or more chosen mediators don't have a KB account.
  • I don't think that the Sysop should limit the pool of KB editors or forum members who would comment on the proposed article versions, since this can also be considered to be part of the mediation process. Gathering comments and ideas from registered forum members may encourage the disputing editors to come to a resolution on their own.
  • If the parties still cannot resolve the dispute, only then should the Sysops as a group select a "arbitration panel" (optionally, the panel may already be in place, as a standing "Arbitration Committee") or another option would be for the disputing parties to choose their own panel to decide the matter.
I know that the above is going into way too much detail and it goes without saying that this is a slow and "expensive" process, in that it takes up a lot of user resources. The most important part of the process is that it should be as free as humanly possible from "political" considerations such as protecting bruised egos, coddling new editors or backing up prolific editors to keep them contributing, currying favor with a forum Moderator or popular forum member (one of the downsides of going to the forums) and the like. Another important consideration is to keep the Sysops out of the actual decision-making process as far as possible, to avoid the criticism that they are abusing their power. Alice Wyman 16:20, 9 December 2006 (UTC)

KB user feedback

The forum thread mentioned above also included comments about quality control and feedback regarding user support documentation in general and in KB articles. I suggested something here awhile back on a feedback process, when problems are found in KB articles (a "bad link" in the specific case I brought up) but got no response. I ended up adding a section on Feedback to the "About" article, but it brings up an important point. Maybe the concept of "KB feedback" should be expanded to include a process by which people on the forums (both helpers and people asking for help) who use a KB article can have a way of giving some feedback on how the article did or didn't help solve a problem? The process should not require the individual to set up a KB account just to post a comment on the article talk page, for obvious reasons. One option is to post a comment to the forum, but maybe a link at the bottom of every KB article to a Feedback form is another possibility (I don't have the technical ability myself to come up with such a page but I'm sure others here do!). Unfortunately, quality control and KB feedback is always going to result in potential alienation of editors if an error they've made is pointed out or if their contributions are rewritten or removed from an article, but that's the price you pay for trying to improve the quality of articles. Alice Wyman 12:00, 7 December 2006 (UTC)

How is that better than using the associated discussion page? The discussion of the tradeoffs of alienating an editor seems to make the assumption there was only one editor for that article. Maybe I have too thin a skin, but if somebody writes an article and does most of the maintenance of it (i.e. they have an emotional investment) I'm concerned about their reaction if it gets publicly criticized and they think its due to changes made by somebody else. The collaborative style of a wiki is the not the type of collaboration most people are used to. I think there is a tradeoff in how much of a big deal you make of any feedback when you don't have many editors and the most common "feedback" nowadays is somebody trying to misuse the article to ask a question that should be asked in the forum. Tanstaafl 15:47, 7 December 2006 (UTC)
There are two types of feedback - that from other editors and that from users. I think the talk pages are fine for editors to post feedback. For users, we really need a "did this help?" feature like many other documentation sites have. Unfortunately, I doubt this is something that "comes with" MediaWiki - it would have to be built.--Np 18:18, 7 December 2006 (UTC)
Don't expect much from users. You will get two kinds of answers: "yes" and "no, it's written for geeks". I don't think you will get much useful information from people who had trouble understanding it, and you are even less likely to get useful technical suggestions. I don't know, you may not even get useful information on what is confusing them. I doubt that it will be worth the trouble.
Also, if they write "no", you might not be able to tell whether it's because the article doesn't apply or whether they just didn't understand. If they don't understand, you can dumb it down, but I think there will be a large fraction of users who won't get it no matter what. I think the best you can do is just write as clearly and as simply as you can, right from the beginning, but keep the information in the article. Feedback won't do anything but waste your time. --AnotherGuest. 8 Dec 06
While I agree that some users won't understand regardless of how it's worded, I still think it would be useful to have feedback on the articles. For example, if 75% of people find one solution helpful, and 30% find another solution helpful, that would say to me that we should talk about the 75% first. If no one found a suggestion helpful, we could remove it.--Np 03:40, 9 December 2006 (UTC)
If you structure it to get what you want, maybe. --AnotherGuest. 9 Dec. 06