User talk:Samuel Blanning/Blanking

From Wikipedia, the free encyclopedia
Jump to: navigation, search

I created this proposal because on WP:AIV I've had to deal with more than one case where an editor was reported for 'blanking vandalism' when the blanking wasn't vandalism, but the reporter obviously thought it was. Problems arise when an inexperienced but well-meaning editor who has never heard of Articles for Deletion, edit summaries and talk pages runs into a Recent Changes patroller prepared to revert what he sees as 'vandalism' automatically, and armed with all sorts of anti-vandal tools that make it easier to do just that (and leave a misfired, impersonal vandalism 'warning' while they're at it). I hope that this will gain consensus to be tagged as a {{guideline}} and linked from the relevant section of Wikipedia:Vandalism. Opinions and merciless editing are welcome etc etc. --Sam Blanning(talk) 19:30, 24 October 2006 (UTC)

Good idea. I've seen quite a lot of this recently. I think though, that this proposed guideline ought to provide some guidence for bot opperators. Most of the instances of non-malicious blanking being reverted and warned that I've seen have involved AntiVandalBot and the like. ~ ONUnicorn (Talk / Contribs) 20:13, 24 October 2006 (UTC)

Mind-reading?[edit]

"If a user believes he has a legitimate reason when removing large amounts of material, he is not committing vandalism." So, um, how are we supposed to determine what J. Random Anon believes when he blanks the day's featured article (without an edit summary, of course—they never leave edit summaries)? Kirill Lokshin 02:55, 25 October 2006 (UTC)

  • Strongly oppose this proposal -- We should not grant users a license to blank large amounts of apparently legitimate text without explanation, and have their edits treated as "content disputes", not vandalism. I would only consider significant content removals to be legitimate edits in the following cases:
1. Where the reason for the removal of the content is readily apparent by examination of the content itself, so no explanation is necessary. In other words, reversions like this should not happen.
2. Where a non-frivolous explanation is provided for the removal of apparently legitimate content in an edit summary. The edit summary should either state the justification for the content removal in its entirety, or provide a link or reference to the page which contains the justification. Note that the explanation for the content removal need not be persuasive to be considered non-frivolous; all that is required is that the explanation avoid substantial absurdity.
In consideration of the fact that new users may be unaware of the need to explain the removal of apparently legitimate content, template:test1a should be used as an initial warning for such removals. However, should the user persist in the unexplained removal of apparently legitimate content, they need to be stopped. The three revert rule is a spectacularly ineffective tool for this purpose, as in situations defined as "content disputes" it applies equally to all parties without regard to the merits of the edits themselves. The three-revert rule can thus prevent a rapid response to inappropriate content removals. Furthermore, the three-revert rule applies on a per-article basis, and thus does not prevent a user from excising apparently legitimate content without explanation from a large number of different articles. John254 02:59, 25 October 2006 (UTC)
I've added two entries to the list of 'obvious vandal/test' edits which should cover the above concerns. The fact is, everyone knows what malicious blanking looks like, so there is no chance that anyone is going to be made unable to stop actual blanking vandalism. However, not everyone seems to know what non-malicious removal of text looks like. This needs to be explained better than it is currently, because it's very difficult to explain it on AIV where things move very fast.
John, the reason WP:3RR doesn't consider the merits of the edits is because the entire point of the electric fence is that if the edits have no merit, more than one person will be willing to revert them, and no-one will need to break it. --Sam Blanning(talk) 03:38, 25 October 2006 (UTC)
Meh, that's probably fine, then. I've also replaced "has a legitimate reason" with "is making a good-faith effort to improve the encyclopedia" (the existing wording from WP:VAND), as I think it's a bit clearer in terms of what those legitimate reasons might be. Kirill Lokshin 05:42, 25 October 2006 (UTC)

Confusing[edit]

This proposal says at the same time that blanking isn't always bad but also that editors should never blank articles. It also makes no mention of blanking for copyright concerns. So what is it trying to say, exactly? And what here is new that's not already in Wikipedia:Vandalism? Fagstein 04:43, 25 October 2006 (UTC)

As the opening section says, "blanking" refers to removal of large sections of text, not just entire articles. And I'm very deliberately saying nothing new - however, what 'blanking' means and what it doesn't is spread disjointedly across 'Types of vandalism' and 'What vandalism is not' on Wikipedia:Vandalism. WP:BLP and some of Jimbo's mailing list posts (such as [1]) are also relevant. This page is aimed at making it coherent. --Sam Blanning(talk) 16:11, 25 October 2006 (UTC)

Some more problems with this proposal[edit]

The set of blanking edits defined as "almost certainly vandalism or testing" and the set of blanking edits defined as "almost certainly not vandalism/testing" are not disjoint. Thus, if "an editor who has a history of good contributions" were to "[blank] numerous unrelated articles", the edits would "almost certainly [be] vandalism or testing" and "almost certainly not [be] vandalism/testing", simultaneously. Of course, this problem could be avoided by replacing the section headings with "Factors that weigh in favor of a finding that blanking is vandalism or testing" and "Factors that weigh against a finding that blanking is vandalism or testing"; however, this would convert the guideline into a difficult to apply, multi-part balancing test. My description above of the cases in which significant content removals are considered to be legitimate edits is far simpler in its application. John254 00:42, 26 October 2006 (UTC)

Carte Blanche?[edit]

Is it just me, or does this read like it gives people carte blanche to blank whatever they want if their edit summary says it was to correct bias or delete false information? I've encountered too many situations where egregiously biased or falsehood-bearing edits were justified as correcting some (imaginary) pre-existing fault in the article to feel comfortable with that. Am I wrong in reading the proposal that way or in anticipating that people will attempt to use it that way? The Literate Engineer 04:10, 28 October 2006 (UTC)

Oppose[edit]

This is either too broad or too narrow; maybe both. It opens the gates of hell for all kinds of trolls to issue forth under the banner of "good blanking". It doesn't sufficiently address the entire issue of knee-jerk, automated reverts of good-faith edits. Let's take this a side at a time.

On the one hand: Generally, deleting large amounts of markup -- text, structural markup, or images -- is a dumb thing to do. The larger the amount taken, the longer the removed material has been on the page, the more likely it is unwise. At some point, it is pretty much mandatory that you simply <!-- comment out --> the portion you wish to remove and speak to the issue on talk. No matter how sure you are that it must all go, it is likely that one or more people put time and trouble into writing the bit and even if most of it is "bad", something can be salvaged.

On the other hand: There has been an explosion of bots and deskbots editing with great speed and little care; in their wake follow editors who see obtuse edit sums and think it's okay to rip through the project with a chainsaw. These "editors" and deskbot drivers confuse quantity with quality and suffer from Editcountitis. There's no difference between a deskbot revert of a sensible deletion and a deskbot revert of a sensible addition; both actions are destructive to whatever congenial community atmosphere we have left, as well as destructive to the project itself.

On the gripping hand: We do not want to create or amplify a category of deletion-as-blanking. This opens the door to wikilawyers who argue this edit or that is not really deletion; it's blanking -- or vise-versa. Remember that the effect of rules is generally to create two new loopholes for every one closed up. There are far too many words on this page: enormous opportunity to process-wank.

I smell here a wide-gauge response to a very narrow situation -- perhaps some one or two incidents that fall within the strained language of When to be especially careful. In my experience, poorly-formed policy proposals with broad scope often spring from small cause. This cause does not look small to the creator; he goes all the way back to first principles, elaborates a new global theorem, and at the end, tweaks and fiddles with it until it finally backs him, retroactively, in a minor matter.

It is wise to exercise good judgement in all we do, to use common sense, to judge, often subjectively, the intent behind others' actions. It is dangerous to allow our judgement to mislead us into hasty action or invite us to bend or rewrite the rules. It is foolish to write policy that requires us to make these judgement calls; somebody will always make an opposing call and then we have a fight.

  • Existing policy is sufficient to handle incidents of large-scale deletion. It's usually vandalism and when it's not, it's almost always unwise. The burden is on the deleting editor to show good faith with a sensible edit summary and clear comment on talk. That showing shifts the burden to the reverting editor even if the deletion is eventually found unwise. Any editor who wants to revert had better leave a better edit sum and a wiser talk. Deskbot output is insufficient.
I note that there are indeed 3RR exceptions relevant here; I wish not. 3RR is "an electric fence". Nobody should cross it, not for vandalism or anything else -- my opinion. You are not the sole defender of the wiki; let somebody else revert that pesky vandal -- it will be his 1st time, not your 4th time. If 3 of your reverts didn't fix the problem, neither will 4. Perhaps a revert from somebody else will carry more weight. Meanwhile, Florida will not slide into the sea.
  • Hasty reverts are often unwise. Frankly, I don't know where to begin to oppose the flood of deskbots and generally disrespectful, hasty editors.
I'm afraid that if things continue as they are, we will move to solve this issue by creating a class of "established editors" -- if we have not already done so with semi-protection and RfA. Established editors enjoy an assumption of good faith in their edits; the burden is upon reverting editors to show cause. The underclass of new and anon editors is then, in contrast, assumed to edit foolishly and in bad faith. If we're going to do this, we may as well do it formally. Close the doors to anon editors, restrict new editors to some sort of mop work, and require registration. Sometimes I think this is wise; most times, I think not.
We surely do need to do something about the misuse of deskbots. We have too many editors who can't really be trusted with a pistol and now they've got automatic weapons.

Drop this proposal with thanks to the creator for raising the issues. Some other action is called for. What? John Reid 14:46, 29 October 2006 (UTC)

Withdrawing as guideline suggestion[edit]

I'm moving this to my userspace and tagging it as an essay; while I think it will be useful to correct someone making obviously mistaken accusations of vandalism, perhaps with some further editing, it's obvious it's never going to be clearly-defined enough to be a guideline. Thanks for your feedback, everyone. --Sam Blanning(talk) 18:53, 29 October 2006 (UTC)