Commons talk:Blocking policy

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

This is the talk page for discussing improvements to Commons:Blocking policy.

Proposal to remove interwiki links[edit]

In 2015 Lotje added several interwiki links (diff). Without digressing into technical details of sysop related processes, this potentially adds a layer of invalid complexity and potential block-related dispute of interpretation. Unless there are good reasons to keep these unqualified interwiki links, I propose to remove them on the basis that they are not suitable "definitions" of these terms for Wikimedia Commons, especially when those definitions may result in, or support, long or indefinite blocks. -- (talk) 10:32, 5 April 2019 (UTC)

Agree that the “w:Wikipedia:Banning policy” link was inappropriate. Few (if any) users were ever explicitly banned from Commons (without all-Wikimedia ban), and anyway Commons currently has no banning policy. Incnis Mrsi (talk) 10:57, 5 April 2019 (UTC)
Thanks, per the diff also to be trimmed are:
Harassment
Sock puppetry
Though these are useful to examine, it would be better to include them as 'see also' material or explicitly qualify them to make it clear that administrators are not bound by the procedures described and the associated links to noticeboards and help pages are unlikely to be correct for incidents of harassment and sock puppetry on Commons. -- (talk) 11:06, 5 April 2019 (UTC)
I support bringing the policy in line with our best current practices, and explaining how that differs from interwiki linked policies on English Wikipedia. However, once we iron out the language, COM:VPP is the way to go, as this policy can affect all users.   — Jeff G. please ping or talk to me 12:08, 5 April 2019 (UTC)

@Steinsplitter: as the admin that protected the page from non-sysops, can you review the above and implement the changes, or remove protection so that I can make them for the community. Thanks -- (talk) 10:13, 7 April 2019 (UTC)

The page is only semi-protected, you should be able to edit it. --Steinsplitter (talk) 10:37, 7 April 2019 (UTC)
Not used to this error type on edit saves. I had to attempt two saves for it to work, and the error message does not actually tell you that is what you are supposed to do, it should probably be worded more helpfully. Not sure if that system behaviour is the same for sysop accounts. -- (talk) 13:07, 7 April 2019 (UTC)

Autoblock disabled when blocking for inappropriate username[edit]

I think autoblock should be disabled when the user is blocked for inappropriate username, because he/she should be able to sign out from his/her user account and continue editing as an IP address. Therefore, I propose the following change to the policy page:

  • Autoblocking of IP addresses used by the blocked user should typically be disabled when blocking bots and inappropriate usernames and but enabled in most other cases.

4nn1l2 (talk) 06:02, 25 April 2019 (UTC)

Symbol support vote.svg Support --jdx Re: 08:56, 25 April 2019 (UTC)

There are usernames and other usernames. For users “Bantu Nation” and “UNESCO office” – yes, disable autoblocking. Not so for “4nn1l2 can't find his dick” and “Copyvio Flood on Abrasive Discs”. The sysop should execute an own judgement. Incnis Mrsi (talk) 14:03, 25 April 2019 (UTC)

Why do you think the word "typically" has been used? Most inappropriate usernames which get blocked are promotional. 4nn1l2 (talk) 21:19, 25 April 2019 (UTC)
Because of haste, perhaps? If the purpose of the block is compelling the user to use another username, then autoblocking of IP addresses used by the blocked user should be disabled. Incnis Mrsi (talk) 07:06, 26 April 2019 (UTC)
That's true and my proposal effectively says so. Most inappropriate usernames which get blocked are promotional. They get blocked to compel the user to use another username (or maybe continue contributing as an IP address). 4nn1l2 (talk) 07:18, 26 April 2019 (UTC)

I applied the change: Special:Diff/349037788. 4nn1l2 (talk) 23:13, 6 May 2019 (UTC)

@4nn1l2: what about "Autoblocking of IP addresses used by the blocked user should typically be enabled for users who were blocked for disruptive behavior. It should be disabled in most other cases, like blocking malfunctioning bots and usernames that don't follow the username policy." - Alexis Jazz ping plz 02:17, 10 May 2019 (UTC)
@Alexis Jazz: I agree with your wording. 4nn1l2 (talk) 01:12, 11 May 2019 (UTC)
No opposition after a week, so I applied your wording: Special:Diff/350833451. 4nn1l2 (talk) 12:20, 17 May 2019 (UTC)

Proposal to change wording for "unauthorized bot accounts"[edit]

Proposal

Current wording:

Unauthorized or non-responsive bot accounts. Bot accounts not authorized by the Commons community are not allowed to operate on Wikimedia Commons, and questionable bot-like editing that cannot be explained by the user should be blocked until discussion takes place. Bot proposals can be discussed at Commons:Bots or Commons:Village pump. Bots may not be operated on Commons without advance permission (which can be sought at Commons:Bots/Requests).

Proposed wording:

Unauthorized bot-like accounts and large scale automation. Disruptive large scale use of automation without a consensus or supporting discussion, including bot-like accounts without bot flags, or large scale changes using tools from main accounts or legitimate sock accounts, may be blocked until discussion takes place. Test runs, bot proposals and proposals for major automation projects can be discussed at Commons:Bots or Commons:Village pump.

This removes the implication that Bureaucrats or any other group needs to "authorize" all non-trivial automation, this is not the current Wikimedia Commons community norm for automation, and it is not a realistic expectation. In particular there is no need to seek pre-authorization for the use of standard tools, or custom automation that does the same job, like cat-a-lot, VFC, AWB, or F2C, so long as the changes being made are not disruptive, but part of the normal housekeeping or other helpful maintenance tasks that may or may not be completed using bot accounts (i.e. accounts with a Bureaucrat approved bot flag). The current wording routinely leads to confusion and delays for high value content projects, with users requesting bot flags for noncontroversial tasks as simple as an upload project or limited and monitored housekeeping tasks, such as recategorizing files for a project with only a few thousand files as its scope.

The new wording introduces the consideration of "disruption", which in practice is the only thing that matters when questions are raised about large scale automation. A good faith user running their own automation projects will always avoid having their account(s) blocked so long as they are prepared to stop the automation while discussion about potential disruption takes place.

Note that malfunctioning bot accounts are under a separate paragraph of BP, so this paragraph only needs to address accounts without bot flags.

Preliminary discussion of this proposal from last month is available at Commons:Bots/Work_requests/Archive_15#Bot_account_misuse,_revisiting_COM:BP. I am quite happy to reword the specific change proposed based on further feedback.

Notifications have been posted at COM:BN and COM:VP. -- (talk) 11:55, 15 May 2019 (UTC)

Symbol support vote.svg Support I also agree it makes sense not to distinguish between bots and tool use. A bot that does good is no worse than someone with cat-a-lot who does good and someone who does bad with VFC is no better than a bot that does bad. - Alexis Jazz ping plz 12:06, 15 May 2019 (UTC)
Symbol oppose vote.svg Oppose There is a difference between blocking a bot-account and a main account that uses automatization tools. The threshold to block a bot-account can be very low (as suggested by the old wording). The same threshold cannot be applied to main accounts that use automatization tools (as sugessted by the new wording, even if they seem disruptive on first glance). A block history on a main account is a different thing than on a separate bot-account. I would be personally offended if my main account ever got blocked (for whatever resason). My bot account got blocked several times already, which I consider a natural part of the bot improvement process. --Schlurcher (talk) 07:29, 16 May 2019 (UTC)
@Schlurcher: Fair enough comment. As there will probably be very few votes on this, so it's more a collegiate consensus rather than a vote, could you recommend a smallish change in the wording that you would prefer? Thanks -- (talk) 16:15, 17 May 2019 (UTC)
Hi, Fæ. I understand these as two different concepts. I rather would like to have two separate statements, i.e. both:
  • Unauthorized or non-responsive bot accounts. Bot accounts not authorized by the Commons community are not allowed to operate on Wikimedia Commons, and questionable bot-like editing that cannot be explained by the user should be blocked until discussion takes place. Bot proposals can be discussed at Commons:Bots or Commons:Village pump. Bots may not be operated on Commons without advance permission (which can be sought at Commons:Bots/Requests).
  • Unauthorized bot-like accounts edits and large scale automation. Disruptive large scale use of automation without a consensus or supporting discussion, including bot-like accounts without bot flags, or large scale changes using tools from main accounts or legitimate sock accounts, may be blocked until discussion takes place. Test runs, bot proposals and proposals for Major automation projects can be discussed at Commons:Bots or Commons:Village pump.
I'm however not convinced that we need the second bullet at all. Disruptive edits (independent if large scale or not) should fall in the category of vandalism and that is already covered. It could be a sub-bullet of that "Disruptive large scale use of automation" --Schlurcher (talk) 07:45, 22 May 2019 (UTC)
Needs improvement. With all due respect to , “disruptive” edits should meet block before growing to become “large-scale”. Yes, I’d like to see administration acting like police. Incnis Mrsi (talk) 17:19, 18 May 2019 (UTC)
What is intended is to fill the gap. We do not block accounts from making annoying but not breaking changes, but if those annoying changes are apparently automated and done at scale, that is systemic disruption and needs to be explicitly blockable without the admin asking for a consensus to get on with it. If you can think of a short and better way of expressing that, please propose some words. Thanks -- (talk) 18:22, 18 May 2019 (UTC)
“Disruptive use of automation at high rate without a consensus, or automated edits persisting after complaints…” Incnis Mrsi (talk) 19:06, 18 May 2019 (UTC)
Thanks, I'll look at integrating it when I'm back from travelling in a couple of days. -- (talk) 19:38, 18 May 2019 (UTC)
Yeah, I could get on board with this. If you are doing a large amount of semi-automated edits, when you see a message on your talk page, you should immediately stop and make sure it isn't someone raising issue with your edits, not carry on for another half our and then see what the message was after 5,000 additional unhelpful edits. In this situation I can see that it would be reasonable to issue a block if the user is non-responsive. But I do still think we should have a general expectation of at least attempted communication prior to blocking, which isn't necessarily the same expectation when it comes to bots. GMGtalk 17:16, 27 May 2019 (UTC)