Wikiversity talk:Blocking policy/Enforcement and self-reversion

Why this was proposed now
These policy clarifications and changes are proposed based on current situations on Wikiversity, as well as experience on Wikipedia. There were two blocks which were matters of concern. This will be lengthy, I may refactor later to summarize and simplify. --Abd 17:26, 21 August 2010 (UTC)

Case:Thekohser
The first was the block of Thekohser, which is complicated. In short, communicating openly through an alternative account which was verified by me to be Thekohser through off-wiki communication, established for this purpose and disclosed to me before any edits were made, and made necessary by a global lock on Thekohser, it was negotiated that Thekohser would make positive edits through IP, self-identify with the edit, and self-revert. Thekohser did so, using 68.87.42.110, and Special:Contributions/69.137.143.91. He would have used only one IP, but a custodian decided that "the experiment was over" and blocked it, even though no disruptive edits had been made, all edits were easily identified, and all were self-reverted. There was only one problem that developed in the trial, and the condition behind it had not been anticipated: what if nobody notices the self-reverted edit? What should the editor do? Thekohser repeated it, which, in hindsight, was an error. Under those conditions, a blocked editor should not repeat such an edit, it unduly complicates history and creates more Recent Changes. Rather, the editor should contact a sympathetic editor on the editor's Talk page, to ask for review, or make a contact off-wiki.

As a result of the trial, which lasted for a substantial time, and the continued cooperation of Thekohser and his refraining from using the opportunity for criticism of the WMF or WMF users or functionaries (which appears to be behind the global lock and his original block here by a privileged user (the privilege was later redefined not to allow such blocks), I unblocked, but the admin then blocked.

The reblock did serve one purpose. As a result, I went to Request custodian action and requested unblock for the account. The discussion determined consensus for unblock, and this was applied to the main account, Thekohser, with a 'crat delinking the account to remove the effect of the global lock. (This is allowed with local consensus, or perhaps at 'crat discretion.)

Had it not been for the self-reversion process, a far more cumbersome method would have been needed to establish cooperation and solicit good content from this user. Ottava Rima proposed it: the user could make contributions on his Talk page, for review and implementation. Thekohser rejected this as too cumbersome, and I agree. An original page might have been established that way, and, in fact, the original page which was created by Thekohser was then blanked, in lieu of reversion -- you can't revert an original page edit --, and then the page was promptly deleted as a blanked page! So probably original pages should be created by blocked users by IP in a different way, to be worked out. Any user could, for example, create a page in their user space for that purpose, the page would be created as a subpage of that, and would be blanked. The user page name would be something like User:Helpful/Pages by blocked editors - temporarily blanked. So that a custodian, seeing the page wouldn't delete it as blanked! The systems must be efficient and easy to follow.

And without a record of good contributions and a record of cooperation with the block -- self reverted edits show explicit cooperation with a block and acknowledge the block and simplify block enforcement -- I would not have had sufficient basis to request unblock for Thekohser, that was my opinion. Some said that I should have gone to meta and requested global unlock. However, that involves dealing with a host of issues based on behavior at other wikis. I was not willing to do that without consensus here, for sure. The chicken and the egg problem. So we created an egg, an account with new life, just a little (I immediately blocked the account, but this allowed Talk page access), incubated the egg and candled it, found it sound, and so we gained, out of a blocked editor, a possibly productive one. And it only takes a single custodian to reblock, if there is new disruptive behavior. I just suggest that any re-blocks require that recusal policy be followed, or be justified by an emergency, actual disruption!

This raised the issue of self-reversion as a process. It raised the issue of an IP block issued without disruptive behavior. My proposed changes deal with both of those issues.

What had been overlooked was that the process of self-reversion, while running, generated good content. It was about demonstration rather than argument. This could have continued for a long time, as long as the blocked user was willing to be patient and someone on-wiki was willing to do the work of review. --Abd 17:26, 21 August 2010 (UTC)

Case:Moulton
The case of Thekohser, blocked in March, 2010, was relatively simple, because there was no local consensus behind the original block. The user was also fully cooperative. It's been said that he was cooperative because he wanted to poke Jimbo by proving that he wasn't truly banned. That could be true, but is actually irrelevant. In order to "make his point," he must remain unblocked here, and in order to remain unblocked, he must not be actually disruptive. Further, to show that he is an active and productive member of this community, he must make positive edits. Generally, when users are blocked on a project, one of the paths back to unblock is a record of positive contributions elsewhere. Hence unblocking him, from his supposedly bad motives, will create positive content. He defeats his own purpose by being disruptive here.

But Moulton is a much older block. It had been discussed to death here, and there was a discussion which, on the face, per a close, banned him. If one reads that discussion, however, there was quite a bit of concern, there was no supermajority, as far as I've seen, for banning. There continues to be agitation on Wikiversity for unblock. Moulton, as well, when the subject was raised with him off-wiki, was quite reluctant to use self-reversion. Why should he? He doesn't care if he's blocked or not. At least that's what he says. In fact, it's inconvenient to have to keep looking for unblocked IP or open proxies. But he's playing the role of Reject (in a game description I've made on Wikipedia Review, and in that game, Rejects have the highest status). Moulton has a possible educational intention in about everything he does. It may be boring for some, and it has certainly been disruptive at times, hence the block. But, then what?

It became clear, watching the IP contributions and the reverts and the IP blocks and range blocks, that the situation was not satisfactory. Moulton would make an edit that was harmless. Sometimes it was on a user Talk page where he was welcome. Jtneill positively welcomed him, and Jt is our most active bureaucrat. So we had the spectacle of an admin reverting welcome edits to the talk page of a bureaucrat, and protecting the page to prevent IP edits!

Yet an admin enforcing a block should not be expected to remember a set of rules about which pages to consider pages where edits of a blocked editor can be reverted, and should not have to examine the edits themselves to determine if they are disruptive or not.

The proposed policy addresses this.

If a blocked editor self-reverts, done. There is no admin action required at all, beyond maybe checking the reversion to make sure it's complete (the set of two edits should result in no changes). If a blocked edit creates a pattern of always self-reverting, it may not be necessary to review every edit. If an editor has been blocked, however, there are usually people who will be trying to catch the editor in some bad act, so it's not very likely that any significant level of reversion failure will happen without notice.

What if a blocked editor makes a decent edit, and does not self-revert? These edits complicate block enforcement, requiring review. Thus the proposed policy allows reversion of them, on sight, by any editor, it need not be an administrator. Whenever I see Moulton edits that have not been reverted, I revert them, unless I know that they are in a place where they are welcome, such as any of several Talk pages. And if they were disruptive, when I was a probationary custodian, I blocked the IP. I did not do this, though, if the edits were not disruptive. That's within custodial discretion.

But if the edits are self-reverted, they are, almost certainly, not disruptive. I've defined a safe harbor: self-identified as being from a blocked editor (I'd use, if I were blocked, "will revert per block of Abd,"), and then self-reverted, should be safe. An IP should never be blocked solely on the basis of such edits, absent egregiously offensive content such that its presence in history is improper, i.e., the edits should be revision deleted. Revision deletion from Moulton edits was happening, to some level, and might continue to happen.

It is not necessary that the self-reversion safe harbor be consider as necessarily leading to unblock. It is not necessary, for the safe harbor to exist at a point in time, that the editor always refrain from true disruption. The concept is that if the editor, editing from some open IP, does not disrupt in any way, which would mean self-reverted edits, self-identified, the community will respect the cooperation for edits from that IP. As soon as the editor starts violating this, the additional block enforcement of blocking the IP can start. Nothing has been lost by allowing a period of contributions, there is simply a return to the status quo ante. The door can remain open, and, in consider all this, it is important to realize that stopping a knowledgeable editor from editing is impossible. We may wish that we could prevent it, but Wikipedia has shown that a determined blocked editor can just keep going as long as the editor cares to, requiring continual attention and continual maintenance. At some points a major chunk of the internet was ranged blocked in a monomaniacal effort to stop Scibaby from creating and editing with socks. To make what were basically harmless edits, quickly reverted (because those articles were "owned," another can of worms. That's how the guy got blocked in the first place, they were owned by administrators. This is dicta, it's not important here. The truth of what I'm saying about it, though, is before ArbComm at the moment, in a massive case on Climate Change.)

We need to simplify this. Self-reversion provides some intermediate state between being unblocked and being banned through enforcement that creates collateral damage. It allows flexibility.

It is possible that a bot could be created to be a "topic ban bot," that would automatically revert the edits of editors under defined conditions; this, again, would establish an intermediate state between "free" and "blocked." Such bots exist on Wikipedia, used as an alternative to the spam blacklist, where there is an oft-spammed site that is also sometimes useful. Certain bot-identified edits by IP are bot reverted. Any registered editor can revert them back in, taking responsibility.

We need to explore ways to move beyond the black and white "free" and "blocked" conditions, the disruption from pure blocking and strict ban enforcement may or may not be as bad as the disruption from the blocked editor, but it is not necessary to accept either condition, we can find agreements if we treat blocked editors with respect. By definition, if an editor has been banned, most of the community doesn't want to continue to deal with him or her. These proposals don't require anyone to pay attention to any edit by that editor, the edits are identified and are, if the editor is cooperating, already reverted. What more could be done? Revert it back in just to argue extensively with it? Only if it raises an important issue! Otherwise, Ignore. --Abd 17:26, 21 August 2010 (UTC)