User talk:Cenarium/Proposal

Comments
In the table, it looks as if, for example, in some situations edits by admins are not immediately visible: is this really what you mean? Surely the right-hand side of the table should be all green, i.e. edits by admins will be immediately visible?

The table is confusing. I suggest separating the table into two tables, current experience and proposed experience. Will semi-protection, for example, be experienced differently under the new system than it is now? This is not clear from the table, or would require careful study of the table to try to figure out what is meant.

We need a proposal where nothing will be more restrictive than it is now: that is, where semi-protection for example will work just as it does now, but semi-flagged protection will allow all the freedoms that semi-protection allows, plus additional freedoms. The table shown here seems to be saying that things will be more restrictive under the new system.

Thanks for putting together some proposals. ☺ Coppertwig (talk) 13:47, 22 February 2009 (UTC)
 * A new revision cannot be automatically reviewed when the previous revision is not, because, for example, an IP could vandalize a section and after that, an admin would edit another section without noticing. If it were automatically reviewed in this case, the vandalism would pass through. Instead, the diff to the latest reviewed revision is shown, thus the admin can see and remove the vandalism, and then review. As for full flag protection, it is used in cases of disputes, so I don't think admins should be singled out.
 * I have divided the table to make current/proposed experience.
 * I have added some related notes. Classic protection levels are not changed, and can still be used temporarily in cases of peaks of vandalism or unstoppable edit wars (and for all non-article pages). I think the second case (an autoconfirmed user edits an unreviewed revision) should be quite rare, as new revisions are continually flagged by reviewers, and so it doesn't affect editing by autoconfirmed users significantly (so it should work like normal semi protection for autoconfirmed users). Cenarium (talk) 14:55, 22 February 2009 (UTC)


 * Thanks for editing the table. I think it's clearer now.
 * Where it says "Can edit; a new edit is visible immediately if the previous version is already reviewed; otherwise[2]not visible to readers by default until reviewed by a 'reviewer'"; I suggest putting this wording only under "autoconfirmed" or something. Under "reviewer" and "administrator", I suggest using different wording (and a different colour: yellowish-green maybe?) with words like "Can edit; a new edit is visible immediately if the previous version was already reviewed or the "review now" option is selected; otherwise it's left unreviewed". (Please verify whether this is an accurate description of what happens.)
 * Otherwise, as I said, your table makes it look as if things will be more restrictive under the new system.
 * I also suggest that in the second table, the first heading be "proposed additional protection levels" if your proposal is to add new levels while also retaining the capability of using actual semi-protection etc. If, on the other hand, your proposal is to replace the old levels with only the new levels, you could put words between the tables like "The above levels will be replaced with the following new levels:". It should be made very clear one way or the other. ☺ Coppertwig (talk) 17:54, 22 February 2009 (UTC)
 * Thanks for your feedback, I'm still working on this. Cenarium (talk) 19:07, 22 February 2009 (UTC)
 * I have updated the table to reflect this. I have no real solution to make it appear less restrictive. The only cases where semi flag protection is less restrictive is when an autoconfirmed user who is not a reviewer edits a page that is not reviewed, as it won't be automatically reviewed. This could happen for pages very frequently edited where new edits come before previous being reviewed. Since they have been semi flag protected for a reason, those pages will often coincide with pages very heavily vandalized and so they may be semi-protected. It may also happen when a page is related to a major current event. When the editing frequency is very high, multiple edit conflicts occur and reviewers will have the same difficulties to review, so in fact it won't be that more restrictive for non-reviewers than for reviewers. Cenarium (talk) 19:01, 24 February 2009 (UTC)

Flag protection

 * Proposal

The major problem here is that reviewers will routinely see something different than the general public. This difference may not be vandalism; it may be clumsy or inaccurate language in the last reviewed version. But it will take longer to fix because fewer people who ar ewilling to edit will see the page. If protecting admins take to reviewing all changes since protection, this will be solved; but will they? 18:19, 22 February 2009 (UTC)
 * All users see the latest revision by default and IPs can decide to see the latest revision. I expect that revisions will be reviewed within minutes as we're not in the case of large-scale implementations. Reviewers will see and flag new revisions at Special:OlReviewedPages. Cenarium (talk) 18:57, 22 February 2009 (UTC)
 * Not what the table says. I very strongly oppose these games of bait and switch. The general public can choose to see what the reviewers see, but they do not have to, and many will not; reviewers similarly will not see what the general public sees, unless they go to the trouble of looking at the last reviewed revision. Septentrionalis PMAnderson 19:30, 22 February 2009 (UTC)
 * What exactly is bait and switch in this ? All registered users see the latest revision by default (not only reviewers) - although they can choose to see the latest reviewed revision by default in their preferences. IPs see the latest reviewed revision by default, but there is a note and a link at the top of the page explaining this, and a "draft" table linking the latest version too. But the time between a new non-autoconfirmed revision and its review should be a few minutes and this would happen only for flag protected pages, so this shouldn't create any problem. Cenarium (talk) 20:30, 22 February 2009 (UTC)
 * A few minutes? Wishful thinking; if we had that many reviewers, we wouldn't even be discussing these proposals. One of the advantages of this proposal is that it is self-limiting; if it stops working, flagged protection will be replaced by semi-protection. Septentrionalis PMAnderson 21:40, 22 February 2009 (UTC)


 * I'd like to point out that, while the system may not encourage anons to edit, in the current system they simply cannot edit (in those situations where they would see something different). { { Nihiltres | talk | log } } 19:03, 22 February 2009 (UTC)
 * I oppose the use of the word "draft" since I believe it contributes to the feeling that one is not being allowed to edit the encyclopedia. The link could be labelled "newest" or  "latest" or something. ☺ Coppertwig (talk) 20:46, 22 February 2009 (UTC)
 * I also don't like this term, but one word should be used as it's for a tab and I don't see what we could use for this (newest or latest alone is not informative). Unless we don't use the draft tab at all, and only leave the short explanation and link at the top right of the page. It would be similar to the protection icons. Cenarium (talk) 17:58, 28 February 2009 (UTC)

Patrolled revisions

 * Proposal

This seems largely harmless. Its chief problem is that patrollers will miss some vandalism, as they do now; this will tend to make such overlooked vandalism survive longer, because anybody who simply checks against the patrolled version will miss it. Septentrionalis PMAnderson 18:11, 22 February 2009 (UTC)
 * We could say the same for reviewing, but does the benefit outweighs the negative ? is the question. It'll allow to significantly improve our oversight of blps without affecting editability. Cenarium (talk) 18:59, 22 February 2009 (UTC)
 * I like this proposal. It's similar to what I've been saying; thanks for giving it a name.  It's analogous to the yellow highlighting in new page patrolling.  The same pros and cons exist there, but I think it's a net benefit.  Those who choose to can still patrol the already-patrolled pages/edits. ☺ Coppertwig (talk) 19:08, 22 February 2009 (UTC)

Deferred revisions

 * Proposal

I don't see that this has much advantage over the present system, where bots simply revert suspect edits, and editors can choose to reinstate them. But it is more complex; is the benefit worth the cost? 18:13, 22 February 2009 (UTC)
 * See Wikipedia talk:Deferred revisions for discussion on this, bots are only used in clear-cut cases (not just suspect edits, I hope so !) and are not always efficient (even for obvious vandalism, see a recent edit to Barack Obama), this could be used for spam, vandalism, test edits, etc. Cenarium (talk) 19:03, 22 February 2009 (UTC)

Technical aspect

 * Section link

Reviewers
What should be the requirements for reviewer rights ?

Requirements for autopromotion can be based on the number of edits, time since registration, number of recent edits, edits in mainspace, ... When should the rights be remove, and when manually added ?
 * Discuss

Rollbackers
Should rollbackers have reviewer rights ?
 * Discuss

Moderators

 * Should we create a moderator usergroup ? Should we grant them other rights ?
 * Discuss