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

Proposal - Modification of Wikipedia Bot Policy to include a mandatory user activatable shutdown facility

The proposal is simple. ALL bots active on Wikipedia MUST have a clear easily accessible facility on their userpage and talkpage enabling ANY editor to immediately shut down the bot if it's causing a problem.

The reasoning is simple. No automatic process should ever be held above a human editor. While many bots already have this facility, it's often limited to administrators only. As many bots are run by administrators, this makes getting a bot stopped a difficult procedure at times.


  1. - The facility must be clear and easily accessible on all relevant pages (bot page, bot talk page, owner page, owner talk page).
  2. - Any user can shut down the bot at any time without fear of sanction as long as they are acting in good faith.
  3. - If the bot is shut down, the bot owner MUST establish communication with the user who shutdown the bot and only reactivate the bot if the user is satisfied that the (perceived) problem will not recur/has been addressed/is a misunderstanding regarding the bot's edits.
  4. - In the event the bot owner and the shutting down user cannot reach agreement, the bot owner must gain community consensus/approval to reactivate the bot.
  5. - If a bot is repeatedly shutdown in good faith, the bot owner may be asked to re-apply for community permission to run the bot again and must show they have addressed the issues/concerns that led to the bot being repeatedly shut down.

Part of the reasoning behind this proposal is that bot actions, even if within policy and useful often aggravate editors due to their inflexible and mechanical nature. Having butted heads with Betacommandbot many times and considered leaving the project due to constant negative attacks on my contributions from it I know firsthand just how discouraging bots can be towards editing here.

In the interests of compromise, I'm willing to amend "any user" to "any autoconfirmed user" to avoid bad faith bot disruptions by anon IP editors/newly registered accounts. Limiting the ability to admins would make this proposal meaningless - any admin can hardblock the bot if they see fit.

Comments and support/opposes below please - well reasoned votes welcome. Abusive or minimalist votes are discouraged.



  1. Vandals would love this feature on ClueBot ... -- Cobi(t|c|b) 21:25, 25 February 2009 (UTC)
  2. For the majority of bots, its incredibly obvious when they break and getting it blocked is as simple as starting a thread on ANI. Many bots run for such short periods of time that they'll finish running for the day/week/month before it gets noticed. The given reasoning makes little sense. "bot actions, even if within policy and useful often aggravate editors" - So a bot can do everything correctly, run within policy, and the vast majority of users like it, and it would still be acceptable for one user to stop it and force a new discussion because they don't like it? Mr.Z-man 21:38, 25 February 2009 (UTC)
  3. Although an emergency shutdown is a good idea, this firstly seems to be a bit instruction creep-y, and very few bots will cause a huge amount of damage before they can be blocked by an admin should they malfunction. I also share Cobi's concern that very useful bots will be kept almost perpetually offline by miscreants once they discover how to disable them. Richard0612 21:39, 25 February 2009 (UTC)
  4. Maybe on some bots.... definately not all. Xclamation point 21:41, 25 February 2009 (UTC)
  5. Yup, I'd shut down addbot immediately and repeatedly. Seriously, an ANI report gets you to the same place with only three or ten more damaged articles. I don't buy the argument that admins run bots => will not shut them down. As I do not see evidence of need for the policy, I oppose it. -Tagishsimon (talk) 21:43, 25 February 2009 (UTC)
  6. If my bot malfunctions I will stop it running. If I am not around then this feature / page we are talking about is already made and is called WP:ANI. This may let the bot edit for a few minutes but removes the risk of people disabling the bots for fun. ·Add§hore· Talk To Me! 21:47, 25 February 2009 (UTC)
  7. Surely there will always be an admin around to use Special:Block. –Juliancolton Tropical Cyclone 21:52, 25 February 2009 (UTC)
  8. Admins on this project are trivial to find 24/7/364 (WE'RE ALLOWED ONE DAY OFF, DAMN IT). --MZMcBride (talk) 22:03, 25 February 2009 (UTC)
  9. Why should we make everyone who opposes the work of a controversial yet consensus supported bot able to stop it? Admins, on the other hand, are qualified, trusted Wikipedians who make decisions like these every day. Let's trust their wisdom on these high-profile bots, shall we? - Jarry1250 (t, c) 22:11, 25 February 2009 (UTC)
  10. This would give vandals too much power. Most problems with bots are quickly solved with a simple note to a bot operator. No need for this; would do more harm then good. Arnoutf (talk) 22:45, 25 February 2009 (UTC)
  11. In the past, it has been said that I have a "fear of bots". This was an incorrect diagnosis; the correct diagnosis would have been a "misunderstanding of bots". I've been doing a lot of reading on the subject, and have learned a lot. And what I've learned is that this proposal will potentially give vandals too much power. Useight (talk) 23:21, 25 February 2009 (UTC)
  12. No. BJTalk 23:27, 25 February 2009 (UTC)
  13. Back when OrphanBot was new, it had this feature. People would trigger the shutdown not because the bot was malfunctioning, but because they disagreed with the policy of deleting unsourced images. FairuseBot has this now, and of the thirteen times the shutoff has been triggered, not one was an actual malfunction -- the most common reason for shutting the bot down is a lack of understanding of the non-free content policy. The second most common reason is illiteracy. I plan to disable the shutoff in the near future. --Carnildo (talk) 23:28, 25 February 2009 (UTC)
  14. Not a chance. A user-operable shutdown function is a good idea for many bots, but this proposal goes far beyond that to the point of absurdity. This seems expressly designed to effectively eliminate all bots from Wikipedia, despite claims to the contrary. Anomie 00:54, 26 February 2009 (UTC)


  • Why is this open for voting before there is discussion? [[Sam Korn]] (smoddy) 21:34, 25 February 2009 (UTC)
    • There's no set way to do this - I simply laid the page out as I saw fit, including the ability to support/oppose from the get go. If votes change during discussion, strikethrough can be used. Exxolon (talk) 21:42, 25 February 2009 (UTC)
  • Part of my reasoning here is the "editor ascendant" philosophy. The editor is king on wikipedia - without them, it would not exist. Bots are not editors and should not be ranked equally or above them. The proposal is one way of ensuring this. Exxolon (talk) 21:43, 25 February 2009 (UTC)
    • So vandals count as editors? Vandals should have more rights than bots? Xclamation point 21:48, 25 February 2009 (UTC)
      • Much as I despise vandalism, there's a principle at stake here - bot's have more abilities/rights than editors as I see it - that is patently wrong. Exxolon (talk) 21:50, 25 February 2009 (UTC)
        • Exxolon, you despise vandalism, but you hate bots. You think that ClueBot has not performed a valuble service here? Xclamation point 21:54, 25 February 2009 (UTC)
          • You seem remarkably quick to make assumptions about me. You've already implied I consider vandals equal to editors. To answer your most recent point it's perfectly possible to deal with vandalism without bots. If we do have to use them, then they should be subject to the restriction I'm proposing. Since I'm willing to have the proposal limited to autoconfirmed editors and most vandalism is from IPs this would not significantly restrict ClueBot from it's task. Exxolon (talk) 21:59, 25 February 2009 (UTC)
            • This is incorrect. A lot of vandals are autoconfirmed (*cough*Grawp*cough*) ... In fact, this experiment was tried with ClueBot ... the page ended up being semi-protected, and then later full-protected due to abuse. -- Cobi(t|c|b) 22:04, 25 February 2009 (UTC)
  • *cough* -- Cobi(t|c|b) 21:45, 25 February 2009 (UTC)
    • I firmly believe that - this is an attempt to compromise between my belief bots should be banned and the usefulness of them that other editors perceive. Exxolon (talk) 21:50, 25 February 2009 (UTC)
  • If your edit is reverted by another editor, that's irritating but dealable with. You can start discussion on the talkpages, get other editors involved and otherwise use the collegial system here to thrash out a solution. If you get reverted by a faceless, mechanical, unaccountable and inflexible bot (which are essentially nothing more than glorified scripts) it's ABSOLUTELY BLOODY INFURIATING - you feel victimised and because you can't discuss things with a bot it's harder to find a solution. Add in unresponsive/dismissive bot operators and you have a recipe for disaster - it drives anyway editors and potential editors who feel their contributions are so unappreciated an automatic process can dismiss them. Exxolon (talk) 21:55, 25 February 2009 (UTC)
    • If bot operators are being unresponsive, they shouldn't be bot operators. If you get reverted by a faceless, mechanical, unaccountable and inflexible bot, you have found a bot that should not be operating. Have you got an example of this? [[Sam Korn]] (smoddy) 22:30, 25 February 2009 (UTC)
      • Bots are by definition all of those things. Since AI bots are not available, they are just scripts and programs, incapable of judgement in their own right. Bot owners vary. I'm using "worst case" example - combine an inflexible bot with an unhelpful operator and you have a damn good way of driving editors away. Betacommand bot, when operating, was a prime example. Exxolon (talk) 01:17, 26 February 2009 (UTC)
        • Certainly bots are faceless and mechanical. They (that is to say, their owners) should not be unaccountable. They can be as flexible as the programmer is skillful! I agree entirely that bot operators should be open, accountable and understanding of others' frustration. I do not see how this proposal will help that. [[Sam Korn]] (smoddy) 01:22, 26 February 2009 (UTC)
  • Exxolon, I used to be like you. Bots angered me because they often beat me to the punch. I would sit down to edit Wikipedia, find some vandalism, click undo, and find out that some bot had jumped in before me. I would get frustrated because I had sat at my computer to edit Wikipedia, not to get beaten and not edit Wikipedia. I tried Huggle and suddenly I wasn't getting beaten endlessly. I could edit when I wanted to. Now my Internet connection is too slow to use Huggle effectively, but while I was using it, I found that I didn't enjoy reverting vandalism enough to do it for even an hour at a shot. I found what I really like to do: improve NFL- and video game- related articles. After reading about the tasks of scores of bots, I realized they do many, many various redundant tasks that are actually extraordinarily useful. If anybody with a four day old account can simply turn these off, the veteran editors will have to use their time turning these bots back on in an endless cycle. And since the veteran editors are a rare sight while vandals are a dime a dozen, this will potentially and most probably, become a huge time-drainer. And manhours are the one thing we are already short on. Useight (talk) 23:36, 25 February 2009 (UTC)

Final Statement[edit]

Since I haven't yet got a single supporting statement, I'm WP:SNOW closing this as a failed proposal. I'll leave you with this question. Can you find a SINGLE article written by a bot on Wikipedia? I'm not talking about a bot that imports things from a database and generates articles based on a template - they do exist. I'm talking about a bot that's written in prose, from scratch, a worthwhile contribution...

That's right. There isn't one. Every single contribution here has been written by a human editor - and yet we are less worthy of consideration than a bunch of if-->then statements. What a joke. Exxolon (talk) 01:21, 26 February 2009 (UTC)

I really don't see how forcing bot operators to jump through hoop after hoop at the whim of any editor solves anything. We contribute to this project as much as anybody else and get annoyed just like everybody else. BJTalk 01:39, 26 February 2009 (UTC)
I can find you loads of articles written by bot operators. I can also tell you that your comparison is largely spurious and unhelpful. No one is ranking bots "above" other editors (last I checked, only an admin can stop you from editing), we are just bowing to the problem of scale. The reason that admins get the 'stop' button is that there are relatively few of us and it takes more than a CAPTCHA script to become one. Otherwise we just end up resetting coren, xlinkbot, clueblot, and so on, each time an editor shuts one of those bots down over a correct application of its rules. Protonk (talk) 03:50, 5 March 2009 (UTC)