Jump to content

Wikipedia:Bots/BetaCommandBot and NFCC 10 c

Add links
From Wikipedia, the free encyclopedia
< Wikipedia:Bots
This is an old revision of this page, as edited by Nandesuka (talk | contribs) at 19:18, 17 February 2008 (→Reply to Lara Love). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

The Debate

As many might know, User:BetacommandBot is causing many discussions in different places, specifically about NFCC10c. Many different arguments are being made and issues are being lost in the noise, and people's issues with the bot are being mis-represented. MickMacNee (talk) 14:21, 17 February 2008 (UTC)[reply]

Purpose of this page

The purpose of this page is to:

  • Bring together relevant previous debates
  • Centralise discussion about how BCB tags for NFCC10c compliance, nothing else
  • State the issues people have with BCB and how it tags for NFCC10c compliance
  • State suggestions for improvement
  • Establish consensus for these suggestions

It should be noted that everything I state here has a disclaimer that I can only state what I have observed or discussions/links I have found or been pointed to. I have requested info before, if none was forthcoming and I have not been able to find it, I make no apology for that. And no, I don't know how to program bots.
MickMacNee (talk) 14:21, 17 February 2008 (UTC)[reply]

Statement of principles

To prevent some of the same mis-representations and mis-directions again, I believe myself and the other concerned editors are:

  • Not interested in changing the policy
  • Not here to advocate copyright infringement
  • Not trying to overule the Foundation
  • Experienced editors who know how to write an FUR
  • Experienced editors who know the NFCC policy
  • Experienced editors who are merely concerned at the effect this bot has on image retention rate
  • Not here for vanity, harrassment of BCB or to prove a point

If any contributors to this page feel they don't agree with the above, there may be better places for your comments, however, I do not wish to stifle opposition to the specific suggestions stated here. If discussions deviate from the stated aims, I or others may redirect them as appropriate.
MickMacNee (talk) 14:21, 17 February 2008 (UTC)[reply]

Statement of issues with BCB

The issues I and others have with BCB tagging for NFCC10c are:

  • The wording of the tag: it is inflammatory, inaccurate and unhelpful to many many editors.
  • The schedule: specifically, the lack of any definition of when it runs, and the act of tagging 15,000-200,000 images in two days, all to be dealt with in 7 days (as the tag states, despite any actual realities). Also the repeated references to some deadline that is to be met, and repeated assertions at the rate of new uploads.
  • The indiscriminate nature of the tags: the bot tags many different situations with the exact same tag, which has different consequences for different people. We recognise that the bot cannot assess an FUR. That is a human's job, but the current blunt instrument approach is clearly inadequate and should be improved.
  • The lack of central information: about this specific tag (NFCC10c), the bot, and the issues surrounding it. This is generating far more debate and taggings than any other issue dealt with by the general message boards or talk pages. This also leads to incredible amounts of wasted effort frequently answering the same questions. This also seems to have produced a general lack of willingness by people who know the answers, to continually answer the same questions this bot produces.
  • The assumption of infallability about this bot and the NFCC10c requirements. It is clear that people think this bot is tagging all non-compliant images, and a bot is needed because humans are too slow. Well, there is a class of non-compliance that the bot never tags, and never will do. Also, the degree of non-compliance issues tagged ranges from the trivial to the severe, all treated the same way with the same tag. FUR's are for a human to judge, by definition.
    MickMacNee (talk) 14:21, 17 February 2008 (UTC)[reply]
  • Demanding more work from volunteers: Wikipedia has volunteer editors. BCB demands extra work from them. If this work is important, and many people support it, those people should be the ones to volunteer. Dragooning editors into doing the work is a hostile attack on the volunteer nature of Wikipedia. I can see the principle behing BCB's operation (they uploaded it, they should fix it), but consider how many editors have announced their resignation from Wikipedia on BCB discussion pages. This should be interpreted as criticism of BCB's operation rather than BCB's mission. RussNelson (talk) 16:51, 17 February 2008 (UTC)[reply]

Proposed solutions

To address the above issues, I propose the folowing actions:

  1. Reprogram or reconstitute BCB for NFCC10c tagging
    1. Separate tagging runs into:
      1. Runs checking newly uploaded images immediately
      2. Runs checking newly uploaded images after a grace period to alow natural human exposure
      3. Runs checking images last changed before the date of a significant NFCC10c change
    2. Separate NFCC10c tagging from BCB into a community bot
    3. Separate NFCC10c tagging from BCB into a BetaCommand operated bot
  1. The NFCC10c tags need to be re-worded (some relate to reprograming above):
    1. State that BCB cannot assess compliance
    2. State that BCB is just a basic check, any issue may be trivial (i.e. moved pages)
    3. State where to find a FAQ board specifically for NFCC10c questions
    4. Remove the 7 day deletion threat if that is not actually the current operation
      1. Make the amount of time before deletion proportional to the amount of time that the image has remained on wikipedia. It's not our job to enforce other people's copyrights. If an image has been on Wikipedia for more than a year and nobody has challenged its copyright status, that alone is evidence of fair use.
    5. State whether an RFU should be for single use or multiple use (i.e. check links)
    6. State whether the tagging is for a new image against the current NFCC policy, or for an image that may now be non-compliant due to a change in NFCC policy
  1. NFCC10c information needs to be centralised.
    1. the reason for BCB needs to be stated, with accurate numeric figures
    2. The pseudocode of BCB needs to be stated, i.e. why it tags an image
    3. The limitations of BCB for assessing compliance needs to be stated
    4. Any planned runs should be stated
    5. Any deadlines being worked towards should be stated
    6. An FAQ needs to be written for the repeated NFCC10c questions
    7. A sub help desk needs to be made and supported for the repeated NFCC10c questions

Feel free to add your own if they meet the purpose and principles of this page as stated above.
MickMacNee (talk) 14:21, 17 February 2008 (UTC)[reply]

Consensus

I hope that people can register their support, opposition or a brief comment on each or all of the proposals in this section. Try to keep the consnsus statements separate from discussions, as per a proposed move type discussion.
MickMacNee (talk) 14:21, 17 February 2008 (UTC)[reply]

Note. Some options are mutually exclusive, i.e. separate the bot.

Support

Threaded discussion will be moved to the discussion section

  • Support most of the above. The details can be worked out as we move forwards. This is a good step to be taking, though I would note that BetacommandBot has been running for nearly two years now (since May 2006), and I think it has been doing image work for around a year. There is a deadline of 23 March 2008, and Betacommand has been working towards that. Give that, this may be a bit late in the day (BetacommandBot has, in the past, tagged tens of thousands of more image than just the ones done in the past week), but it would be good to get agreement on how to handle things from now on. Carcharoth (talk) 14:37, 17 February 2008 (UTC)[reply]
  • Support Will (talk) 14:53, 17 February 2008 (UTC)[reply]
  • Support. My specific objection is that BetaCommandBot makes nearly-constant mistakes, incorrectly tagging images that have a valid fair use rationale. This is not about opposing the policy. The policy is fine; the issue is that BetaCommandBot is doing its job poorly.Nandesuka (talk) 14:59, 17 February 2008 (UTC)[reply]
  • Support. I notice that Hammersoft unconditionally supports the current operation of BCB. That kind of blind support for something which is clearly flawed is reason alone to support a call for changes in BCB operation. RussNelson (talk) 16:56, 17 February 2008 (UTC)[reply]
  • Support. We need these changes to the way the bot handles 10c. I do not object to the fair use policy, just to the way this bot applies one particular rule of it. Those who are saying "you can't change the bot! it's policy!" need to realize that a different, better bot could also work with the policy without being so harsh, inflexible, and disruptive. rspeer / ɹəədsɹ 19:15, 17 February 2008 (UTC)[reply]

Oppose

Threaded discussion will be moved to the discussion section

  • Oppose: Just another lynch mob for the bot, which was approved by WP:BAG, approved by the bureaucrats, supported by Foundation resolution, and supported by local EDP policy. This page reeks of bias against the bot. If you want to solve problems, wipe this ridiculous slate clean and start from an unbiased approach. --Hammersoft (talk) 15:52, 17 February 2008 (UTC)[reply]
  • Oppose - I agree with Hammersoft. This needs to be deleted and redone without bias. Some of the comments made in this proposal are completely flawed. We are all volunteers, this is true. However, it is not an unfair burden to place on the uploaders of images to bring those images into compliance. To demand all of this extra work on Betacommand to recode his bot for all these things is absurd. Wikipedia is the free encyclopedia. Those who desire to bring non-free content into it carry the burden to do so within policy. Editors who are too stubborn or too lazy to do this and choose instead to leave the project should not be held in high regard. That aside, a copyrighted image remaining on Wikipedia for a year is not evidence of fair use. I'd like to see someone put "This image is allowed under fair use because it's been sitting on Wikipedia for a year unchallenged" and see how that flows when we get hit with copyright infringement. Adjusting the grace period from tag to deletion to be proportionate to how long the image has been on WP is just ridiculous. The fact that the March deadline is being questioned is laughable. While there may be merit to points in this proposal, overall it's just more bitching about a bot that annoys some people but does overwhelmingly good work that would otherwise be unmanageable. Lara❤Love 17:31, 17 February 2008 (UTC)[reply]
  • Oppose It is a fundamental principle of the project that Wikipedia should be free content. Unfree images which do not meet with WP:NFCC should be removed or fixed to meet the policy as soon as is possible. The bot appears to identify non-compliant images correctly in almost all cases and at that point all responsibility for "saving" the image from deletion lies with any users who wish to do so. Guest9999 (talk) 18:22, 17 February 2008 (UTC)[reply]

Discussion

  • I thought that WP:ICHD was specifically created, or possibly repurposed, to address BCB issues? Mind you, this page should be linked into Beta's warning, and the top needs to have a brief FAQ for dealing with BCB requests and the limitations of BCB. Also, I don't know if Beta is looking to or is able to program the bot to group warnings to a user (a very common complaint when it hits up 1000s of images at a time), though some of his newer edits suggest he might be looking into that; however, we should mention in the BCB limitations that it is unfortunate that BCB is spamming talk pages when it does large runs, but this is a necessary side effect and attempts to mitigate it are being looked at. --MASEM 15:23, 17 February 2008 (UTC)[reply]
Spamming is not necessarily unwarranted, some users upload many pictures. But with the scheduling and separation of run types would stop all the images you might have uploaded in 3 years being tagged at once. MickMacNee (talk) 15:26, 17 February 2008 (UTC)[reply]
  • Discussion of Hammersoft's oppose:
  • One, stop refactoring my comments. Two, my answers are extremely specific. This is a detestable page that is heavily biased. It is inherently flawed in its language, premise, and approach. I made very direct recommendations on what should happen to this page. Perhaps you would be so kind as to exert some energy and fix the blatant bias in this page? --Hammersoft (talk) 16:10, 17 February 2008 (UTC)[reply]
Hammersoft, "approved by WP:BAG, approved by the bureaucrats, supported by Foundation resolution, and supported by local EDP policy" - I agree with some of this, but I think you are overstating the case. Please provide diffs for something where the Foundation specifically supported the way BetacommandBot operates? And remember that some people reading this page won't understand what EDP means. Please state "Exemption Doctrine Policy" and explain what that is, if you are going to use that term. Carcharoth (talk) 16:37, 17 February 2008 (UTC)[reply]
Hammersoft, I think that you would call any listing of problems with BCB "biased". Yet I was not biased against BCB until it spammed a page in my watchlist. I had no opinion (was unbiased) until I saw it in operation. Instant hostility on my part, and .... I'm not alone. BCB angers MANY people. Some of them are made angry simply because they are being asked to do extra work (work that they should have done in the first place). Some of them are made angry because of the manner in which BCB informs them of the problem.
  1. It's text comes across as hostile, and I've suggested improvements which have been ignored.
  2. It imposes a 7 day deletion time which is actually not enforced, so it's lying.
  3. The 7 day deletion time is completely arbitrary and has no relationship to the amount of time the image has been on wikipedia without being challenged by the copyright holder.
  4. BCB doesn't make any use of the number of pages to which an image is linked. If an image is only used in one place, it should insert a different notification than if the image is used in multiple places.
  5. It inserts MANY notifications when one is sufficient.
How can you continue to defend BCB in light of all of these flaws? I'm guessing that you're a friend of Betacommand. I understand loyalty to one's friends. However logic demands support for changes to BCB, and right now you're not being logical. RussNelson (talk) 17:27, 17 February 2008 (UTC)[reply]
7 days is a long agreeded standard as a balance between giving time to the uploader and not leaveing things hanging around forever. Wikipedia always has backlogs which means that at any time some of our tags will be in error. This is a feature not a bug since it means we don't constantly have to update the tags based on backlog length.Geni 17:54, 17 February 2008 (UTC)[reply]

At this point with only a little over a month left to run before the foundation policy is finalised we have little choice but to use BCB style tactics.Geni 17:54, 17 February 2008 (UTC)[reply]

Russ, this is an extreme example of bias. It's also extremely rude. You're calling those who disagree with this preposterous proposal illogical and saying it's only as loyalty to a friend. Seriously? Negative. The proposal is illogical. The text of the tags and notices are not hostile, they're to the point. The bot does not lie, the writers of this proposal are misrepresenting. The bot does not say that the image will be deleted in seven days. It says after seven days. So whether it gets deleted next week or in June 2012, the bot hasn't lied. And even if some of the tags said "in seven days", the fact that admins can't keep up does not make the bot a liar. Be glad your grace period has been extended and stop begging for an extending grace period. And to expect the bot or anyone else to base the grace period on the amount of time the image has been on Wiki... what? Whether the image is used in multiple articles or one, the point is that the FUR needs to be specific to each use. A change in wording is arbitrary, not crucial. And last, one notice is not sufficient. If I've lazily uploaded a dozen copy-vios and the bot only notifies me of one, then I'm only going to fix one... or, as is apparently shown historically, go pout about it and leave the project. Either way, there needs to be a notice for each image. If your talk page gets smacked with a dozen notices, take it as a clue-by-four to the face that you need to reevaluate the quality of your uploads. If they are historical uploads, then I guess it's time to go ask some friends for help if you can't be bothered to fix a handful of images yourself. If the uploader has left, then if the images are so necessary, I'm sure they'll easily be reuploaded... hopefully appropriately. Lara❤Love 18:29, 17 February 2008 (UTC)[reply]
A couple of points. A brief look at BCB talk should be sufficient to see how many people are confused about being asked to provide an FUR for all instances when there is only one instance, and due to the way the image page works, it has a link, and either by looking at the guidline or due to legacy NFCC changes (again not differentiated by the bot) the image has what looks like an FUR. Also, re-uploading is not that easy is you don't know where it came from, again a possible bot improvement, a subsection of images tagged where the uploader has not editted for x months and unlikely to see the tag. And finally, even the statement will be deleted is innacurate, whether in or after 7 days. No one can say for sure that a tagged image is definitely going to be summarily deleted when examind by a human, which is the only way compliance can be judged (again as said above, this has never been assessed), especially as complaints are continually treated individually. MickMacNee (talk) 19:03, 17 February 2008 (UTC)[reply]

Reply to Lara Love

To reiterate what I tried to make clear on this page, this in not about lazy/inexperienced editors or first time uploaders, and not about challenging changing the policy. Try and read the specific points from that point of view. Also, the comments about volunteers is not totally my point, but it is relevant to the specific schedule that BCB is being run to, i.e. 20,000 images in 2 days and then nothing. The point about loading work onto betacommand, he can solve that by releasing his NFCC code (should he be allowed to ignore the multitude of requests for changes just because he coded it for free?), or he can take what I imagine would be a couple of days works to do as suggested and split tagged images into different categories, to reflect the very different types of non-complance that it catches with one tag. That would at least allow the uninvolved people who wish to save images, to at least pick the ones that take 10 seconds to fix, but as it was uploaded in 2005 and obscured in a huge catch all net, doesn't bother. And bear in mind, it increasingly seems as if hardly any of these tagged images are ever not saveable on a point of law, and once again, BCB is NOT assessing comliance with the law, as pointed out above. MickMacNee (talk) 18:08, 17 February 2008 (UTC)[reply]

If the image does not have an adequate fair use rationale, then it's not in compliance with policy or law, correct? Is the policy not tailored to the law? Even if the policy is more strict than the law, that's an issue to be taken up with the policy, not the bot enforcing it. And if these images are so precious and critical to the encyclopedia, why spend so much time on complaining about it instead of working to save them? For all the attention being drawn to BC and his bot, a drive could have been started to bring in the help of others to save all these historically necessary images. Lara❤Love 18:38, 17 February 2008 (UTC)[reply]
You are confusing the policy with the bot, as I thought I had adequately explained above. I was saving images, until it became clear that it was pot luck whether all an image required was 10 characters added, or a whole rationale, or was infact illegal and not allowed at all (and I have absolutely not encountered a single case of this as yet); and also that it was pot luck as to how many turned up in my watchlist to save, none, none, or 8, or none. So no, blindly getting on with fixing tagged images as and when does not seem an efficient use of time to me considering how I observe that BCB is currently being run or has been coded. And instead of making continued compliants, here I am, hopefully not wasting any more time that I realy wish I was using to edit articles, but on the 20,000 run day, I didn't get much of that done I can tell you, and I have a very small watchlist, and again, none of these tagged images are my uploads. MickMacNee (talk) 18:52, 17 February 2008 (UTC)[reply]
How am I confusing the policy with the bot? Repeating something doesn't make it true. The bot enforces policy. Period. If you don't like that policy, go try to change it. Lara❤Love 19:04, 17 February 2008 (UTC)[reply]
The bot enforces policy, but the bot and policy are not equivalent. Better bots could enforce policy too. We want to change the bot, not the policy. rspeer / ɹəədsɹ 19:17, 17 February 2008 (UTC)[reply]
The policy is that all fair use images must have a valid fair-use rationale. BetaCommandBot (a) tags images that do have a valid fair use rationale, and (b) doesn't tag many images that don't have a valid fair use rationale. The policy is fine. The problem is that the bot is doing a bad job of applying the policy. That's why we're here. Nandesuka (talk) 19:18, 17 February 2008 (UTC)[reply]

Reply to Guest999

"The bot appears to identify non-compliant images correctly in almost all cases". First of all, one of the issues stated above is that there has been no demonstration of this with public analysis. Second, there is a large sub-set of illegal images that would not be tagged by the bot and never would, thereby almost all cases is incorrect (this could be confirmed if the code was explained - another suggestion made above). Third, this comment reinforces my view stated here that many people are mistakenly believing that this bot is making an objective assesment of compliance, it is not, it is making a crude string match, nothing more. This bot is a helper, it should be seen, described and treated as such, both in the tag and elsewhere, before anyone faced with it is given that impression. It absolutely does not and never can assess comliance to any meaningful degree, and in the cases of images with single use, it is coming up very short of even identifying images that would not pass muster in a court as having a valid rationale, due to which version of NFCC was in force at the time. Again there is a suggestion that would help differentiate these cases for easier fixing, by better categorisation and more helpful wording, because at the end of the day, this is about fixing things. You cannot say that the wording of tags has no effect on the likelihood of things getting fixed. Finally, I am very concerned at the idea that enforcement of fundemental principles of WP should be turned over to bots. What next, civility bot? Banned for 24 hours if you swear? MickMacNee (talk) 18:44, 17 February 2008 (UTC)[reply]

  • First, I am amused by the reasoning behind RussNelson's opinion. I understand that showing support for a bot is a very scary idea which must be oppressed, and that the only opinion that can rightfully be had about bots is that they are evil and out to destroy the wiki. However, we seem to have quite a few of these bots running, and the support of a bot by one user is not a valid reason to oppose it. Please find a more cabalistic and uninformed reason and return. Now, your proposals. I see that you have numbered them. I see that they are all numbered 1. This attempt to confuse me shall not be taken lightly. Nevertheless, I shall evade this assault and use the words 'first', 'second', and 'third'. Regarding the first proposal, subsection 2. What is a 'community bot'? One whose code is changed based on consensus? This is truly a bad idea. People oppose BCBot and others like it because they disagree with the policy it enforces (or don't understand it) and by being allowed to change it, would alter its purpose. This is a bad thing.
    "It's not our job to enforce other people's copyrights."
    Yes it is. The foundation said so.
    "If an image has been on Wikipedia for more than a year and nobody has challenged its copyright status, that alone is evidence of fair use."
    So why don't we protect all articles that were created before February 2007? People have had plenty of time to change them! In fact, since we've had plenty of time to make all the new articles we need, let's close article creation. Heck, let's close the whole wiki down and publish! It's been enough time! We've fixed all the problems by now!!
    "Dragooning editors into doing the work is a hostile attack on the volunteer nature of Wikipedia."
    Any editor who thinks the image is valid is free to correct it. We only notify the uploader because they are the ones who know about it. If you'd like to help, there are categories of these images.
    "State whether an RFU should be for single use or multiple use (i.e. check links)"
    What is an RFU?
    "the reason for BCB needs to be stated, with accurate numeric figures"
    Why? BCB is needed because images exist which violate policy. Why do we need numbers of that? Why must we waste BC's time in getting those numbers?
    "The pseudocode of BCB needs to be stated, i.e. why it tags an image"
    Again, why? If you think that a tagging was done in error, then tell BC, and he can fix it. There's no need to give out the regexps and algorithms, because then he'll be forced to justify every bit of it, even where a change was made for the better. Here we are with giving BC more work to do. And we whine about how the bot screws up so much.
    "Any planned runs should be stated"
    Why? I run my bots on a cronjob or when I get a chance to get online. I don't see a need to post this information, especially when it's all available in Special:Contributions.
    "A sub help desk needs to be made and supported for the repeated NFCC10c questions"
    I believe there is a link to one from BCBot's talk page, STBotI's talk page, maybe even BC's talk page.
    "An FAQ needs to be written for the repeated NFCC10c questions"
    How about that big box on the top of User talk:Betacommandbot that no one reads? Why would they go out of their way to read a FAQ when they don't even look at the big red stop sign?
    --uǝʌǝsʎʇɹnoɟʇs(st47) 18:48, 17 February 2008 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Bots/BetaCommandBot_and_NFCC_10_c&oldid=192115317"