Wikipedia:Bots/Requests for approval/SSTbot 2

The following discussion is an archived debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA. The result of the discussion was

Withdrawn by operator.

SSTbot 2

Operator: SSTflyer (talk · contribs · SUL · edit count · logs · page moves · block log · rights log · ANI search)

Time filed: 08:30, Monday, August 8, 2016 (UTC)

Automatic, Supervised, or Manual: Manual

Programming language(s): c:Help:Gadget-Cat-a-lot

Source code available: User:SSTbot/common.js

Function overview: Performs category-related tasks as required when closing WP:CFD discussions

Links to relevant discussions (where appropriate):

Edit period(s): When required

Estimated number of pages affected: Depending on CfD discussion

Exclusion compliant (Yes/No): No

Already has a bot flag (Yes/No): Yes

Function details: As a non-administrator, I am not allowed to use WP:CFDW when closing CFD discussions. Currently I use Cat-a-lot to rename, merge or remove categories on articles, but several users have expressed the concern that such edits overflow their watchlists. I am asking for the permission to route such edits (which I already perform routinely) through a bot account, so that users are able to hide them from their watchlists.

Discussion

Aren't there many bots already doing this? I'm more concerned that you will only be partially completing the activity as you won't actually get the category deleted. — xaosflux ^Talk 23:20, 8 August 2016 (UTC)[reply]

In many cases (such as merge or rename closures) the original category page does not need to be actually deleted. Like all non-admin closures, I do not have the ability to delete pages. I already routinely use Cat-a-lot. This is a request for such edits, edits I currently perform with the User:SSTflyer account, to be routed through a bot account. It's not automated. SST flyer 02:27, 9 August 2016 (UTC)[reply]

I think Plastikspork's Sporkbot does this already and in an almost automated way. -- Magioladitis (talk) 07:09, 9 August 2016 (UTC)[reply]

My bot orphans templates after TFDs are closed. I don't do any CFD stuff. Thanks! Plastikspork _―Œ^(talk) 01:13, 10 August 2016 (UTC)[reply]

Aren't these bots already handling CFD sufficiently (and with delete powers too) - it looks like one is suspended as not even needed. — xaosflux ^Talk 20:19, 16 August 2016 (UTC)[reply]

Yes Xaosflux, but such bots can't be used by non-admins, and so the only way they can close merge and rename discussions there is by either changing the category manually, which is tedious, or by using a tool like cat-a-lot, which floods watchlists. Omni Flames (talk) 22:57, 16 August 2016 (UTC)[reply]

Is that "Yes also" to the "handling CFD sufficiently" part? I'll invite some CfD regulars over to this discussion. — xaosflux ^Talk 23:48, 16 August 2016 (UTC)[reply]

@Xaosflux: Thank you for the ping. Currently, the only way to have the existing bots process a category is for an admin to list it at the fully-protected WP:CFD/W. The full protection has been there for a long time, minus a brief abuse of template protection when I was a non-admin template editor who closed CfD discussions (not requested by me, just for the record). It's necessary given the potential for massive vandalism from these bots if something like Category:1990 births was plugged in as "orphan and delete", but this is a significant inconvenience to our non-admin closers, who have to process the tedious orphaning and merging themselves. I nearly requested a similar task as a non-admin and would certainly have done so if my RfA were not successful. The "need" for a bot here lies in our need for additional active non-admin closers given the massive CfD backlog. This task is very necessary, very useful, and well within the bounds of existing consensus at CfD if SSTflyer intends to actively close CfDs. In fact, I strongly encourage SSTFlyer to bump this up to blanket approval for an automated task so he can focus on the actual closing. I wrote AWB scripts as a non-admin which are functionally equivalent to the various tasks the CFD/W bots perform, and I'm happy to provide them to a long-term positive contributor. ~ Rob₁₃^Talk 00:11, 17 August 2016 (UTC)[reply]

Not a good idea. The solution is simple: non-admin closers should close discussions only when they have the tools to implement that closure. This is just an attempt to use technical means to make an end-run around that long-standing principle.
If SSTflyer wants admin tools, then WP:RFA is the place to get them. --BrownHairedGirl (talk) • (contribs) 07:47, 17 August 2016 (UTC)[reply]
- Administrator tools have nothing to do with orphaning categories, which is the next step after a close as delete. Non-administrators do have the technical tools to implement the next step of any CfD close. ~ Rob₁₃^Talk 14:52, 17 August 2016 (UTC)[reply]
Until/unless the policy is explicitly changed to allow non-admins to close CFD discussions (I would be opposed to such a proposal, FYI, although I may be willing to allow template-editors to close renames and deletes per the technical protection issues), these discussions should only be closed by admins except in blatant speedy keep situations (including withdrawn nominations). עוד מישהו Od Mishehu 08:29, 17 August 2016 (UTC)[reply]
- This has already happened. See the closing instructions for CfD, which allows non-admin closures. That's been the practice for years. This is not new, and BRFA is not the place to change our closing instructions to exclude non-admins. (As an aside, if I hadn't closed CfDs as a non-admin, we'd have a backlog going back to April right now.) ~ Rob₁₃^Talk 14:51, 17 August 2016 (UTC)[reply]
  - U, hold on Rob. I have just done some checking.
    The closing instructions restricted non-admins to no-change decisions until this edit on 23 July 2015 by Fayenatic london. That is not "years", as you incorrectly assert; it is less than a month over one year.
    I am sure that FL was acting in good faith in response to technical changes, but I am not aware of any consensus-forming discussion which agreed this significant change. The fact that non-admins gained the technical ability to do some of the tasks does amount to a consensus that it is appropriate for them to do so.
    CFD closures can affect many hundreds of articles, with a set of changes which are hard to track, let alone to revert. For that reason, I believe that it is preferable that non-keep closures should be a ) made by an admin, whose judgement has been assessed by the community, and b) implemented by a regular bot which uses consistent edit summaries to assist monitoring and tracing. (The latter reasons makes CataLOt a dreadful tool for CFD closures, because it provides no explanation. There should be a ban on using it for CFD closures).
    I suggest that there should be an RFC on CFD closures, rather than yet another stealthy change to long-standing practises. --BrownHairedGirl (talk) • (contribs) 21:23, 17 August 2016 (UTC)[reply]
    - While that may have been the date of the edit, I've been told it was standard practice for years prior, something you may wish to confirm with FL. This massive change conducted at a BRFA would be the true stealthy change. WP:NACD is the actual guideline governing non-admin closures of deletion discussions. It states "Non-administrators should limit their closes to outcomes they have the technical ability to implement", and non-admins do indeed have the technical ability to orphan categories, which is the next step in the deletion process. Non-admins can close TfDs as delete for exactly the same reason, and there was a large-scale RfC confirming that interpretation of the guideline.
      I have little doubt this BRFA will fail due to justifiably conservative attitudes from BAG members, who would probably be right to deny this to avoid controversy given the unexpected opposition to a practice that has existed unopposed and well-documented for a year (and not-as-well-documented for longer). This will then be brought to the community, and the community will probably reject the idea of non-administrators closing CfDs as delete based on a complete lack of understanding of the available resources at CfD, the historical quality of non-admin closes there, and the active damage done to the project by a large CfD backlog. Simultaneously, the current non-admin closers won't be able to pass RfA due to the absurd standards that have materialized there, whereby a lack of GA-level or higher content work is equivalent to being a power-hungry monster who's here only for personal gratification. I'll be happy if I'm wrong about any aspect of that, but if I'm not, I'm walking away from CfD entirely. It's an exercise in futility to beat our heads against the backlog while the community systematically mugs us of our tools to actually reduce or eliminate the backlog in a meaningful way. Fayenatic london and Good Olfactory, you're welcome to comment. I need to step back. ~ Rob₁₃^Talk 21:48, 17 August 2016 (UTC)[reply]
      - Rob I respect your right to disagree, but I do wish that you would state your case with a little less hyperbole, and a little more checking.
        I have been editing here for a lot longer than you, and I am not aware of any significant number of NAC CFD closes until this calendar year. Rather than "I am told" assertions, it would have been wiser for you to do some checking of the basis for your assertion. I didn't just rely on memory; I spent a few minutes using AWB to scan all of the CFD log pages for 2014, and I found only 4 pages which include the text non-admin closure, as required by WP:NACD. In each for the 4 cases (2014 April 15, 2014 August 21, 2014 January 2, 2014 May 9) there was only one NAC per page, and in each case it was a speedy keep.
        So if CFD had a lot of NACs back then, they were being done stealthily, contrary to the guidelines ... and that's no basis for claiming that there is a consensus to continue this way. --BrownHairedGirl (talk) • (contribs) 23:25, 17 August 2016 (UTC)[reply]
        I don't think it's hyperbole to state that we have a large CfD backlog, a means to clear it, and aspects of the community who constantly seek to take away that means. I know it's not hyperbole to say that I have no intention of working on backlogs which the community actively fights against having cleared. There's always another backlog out there, and I'm happy to switch to something else. I do respect your experience editing here, but I also respect the fact that until August 2015, there was no requirement to state you were an admin when making closes, just a recommendation. (Interestingly, this bold change occurred more recently than the change to CFDAI.) ~ Rob₁₃^Talk 02:06, 18 August 2016 (UTC)[reply]
        Rob, it would be a god idea to more carefully consider the texts you are citing.
        The version you link to says "Non-administrators closing deletion discussions should state this fact in the closing decision".
        The current version says "Non-admins should indicate their non-admin status with the {{nac}} ("non-admin close") template".
        So the advice was there, and has not been strengthened. All that's changed is the specific mention of the template.
        All of which brings us back to the point that you have still offered no evidence at all for your secondhand assertion that NACs at CFD was standard practice for years prior. --BrownHairedGirl (talk) • (contribs) 09:01, 18 August 2016 (UTC)[reply]
        @BrownHairedGirl: Whoops, I linked Special:PermaLink/674442285 when I meant Special:Diff/674442285. Sorry for the confusion. ~ Rob₁₃^Talk 09:09, 18 August 2016 (UTC)[reply]
        Rob, that diff[1] shows a change from "recommended" to "should", which is not a huge change. There is a slight strengthening, but neither wording is mandatory.
        The point remains that neither my memory nor any checks I have done support your assertion that non-admin closures were a long-term feature of CFD. Your angry hyperbole about BRFA being abused to overturn a long-term practice remains unsupported by any evidence other than gossip, which you have apparently made no effort to verify. --BrownHairedGirl (talk) • (contribs) 09:43, 18 August 2016 (UTC)[reply]
        You're quite right that I can't back up the claim and so can't assert it; I mentioned it a single time as an assertion, mentioned where I had obtained the information when challenged, and then dropped it since I do not have the time to dig through ancient history to back it up. That does not change that this BRFA is being used to change our practices. WP:PGBOLD, a policy, allows bold editing of policies and guidelines. Closing instructions are neither, so they certainly may be edited under the same terms. As per WP:EDITCONSENSUS, a policy, "Any edit that is not disputed or reverted by another editor can be assumed to have consensus. Should that edit later be revised by another editor without dispute, it can be assumed that a new consensus has been reached." Given that we've gone over a year without challenge to those closing instructions or the fact that myself, Marcocapelle, and others have made hundreds of closes under the new instructions, it's hard to believe we still aren't at the point of assuming a new consensus has been reached. I brought hundreds of non-admin closes of CfDs (including the use of WP:CFD/W) to my RfA, and not even my harshest critics suggested that non-admins shouldn't be allowed to make such closes, providing evidence that the community doesn't consider this controversial. I've already noted that the community has endorsed at a large-scale RfC the notion that non-admins may close discussions where the next step in the process is not deletion. I don't know what more I could possibly say. It's possible I was incorrect to use the term "years", but that is hardly central to the argument that the current practice at CfD is to allow non-admin closes. ~ Rob₁₃^Talk 10:10, 18 August 2016 (UTC)[reply]
        It may not be central, Rob, but it is highly relevant.
        The long-term practise at CFD was admin-only closures. AFAICS, that persisted until earlier this year, and I am not aware of any RFC agreeing to change that long-established practice. Many of the subsequent non-admin closes have been made without the {{subst:nac}} tag, so they are not so readily identifiable and thus less open to scrutiny.
        I don't see any sign of any intent at subterfuge, but the effect of a set of poorly advertised cumulative changes amounts to a stealthy shift which can hardly be described as a consensus.
        I also disagree with your comments about the quality of the NAC closures. I have seen a significant number which I felt were dodgy, and regret not taking the time to challenge more of them. In particular, I note a persistent trend among more than one of the newer closers to fail to identify that a nomination proposes merging an intersection category to only one of its parents, without any rationale for the failure to double-merge. (e.g. if Category:Fooian XXXists is considered too narrow, it should usually be upmerged to both Category:Fooian people and Category:XXXists, unless there is some special reason not to. However, I see some of the new closers repeatedly failing to identify this either in their own nominations or in their closes. If this sort of thing is not spotted, it can have devastating effects to the integrity of the category tree.) --BrownHairedGirl (talk) • (contribs) 14:13, 18 August 2016 (UTC)[reply]
        Very good point, BHG. I suggest we incorporate it into CFDAI. At the moment, if no one spots this pattern during a discussion, the consensus is to merge to a single parent, and the closer may naturally do so. My own practice as a bold admin is to merge to both parents, and to state in my close that I am doing so because the discussion has not given a justification for removing the member pages from one of the hierarchies. However, other admins closing CfDs – and not only CfD newbies – may overlook this and feel justified in following the written consensus. Indeed, in less obvious cases than Fooian XXXists, the second hierarchy would only be noticed when taking a look at the nominated category. – Fayenatic London 08:21, 22 August 2016 (UTC)[reply]
Comment. It seems to me that User:SSTflyer is in a bit of a catch-22 here. The user wants to assist with CfD backlog, but doesn't have access to the tools. One way he could get access to the tools through an RfA. However, the RfA process has become (in my view) tremendously demeaning to users in the way that existing admins there challenge the motivations and intentions of applicants. I nominated User:BU Rob13 for adminship, since he was largely in the same position as SSTflyer w.r.t. wanting to assist with CfD, and in retrospect I can't believe some of the objections at his RfA and what he was asked to "answer for". I think he too would agree that in retrospect, he would not have bothered since the trade-off was hardly worth it. So perhaps wanting to avoid all of that, SSTflyer tries to get a bot that he can operate to allow him to do some of things needed to close CfDs, and he is just told that it's a bad idea and that he should use RfA. The CfD backlog will never go away if users continue to get stuck between this "rock and a hard place". Personally, I think we should just experiment with unprotecting WP:CFDW and see how it goes. I could agree that if vandalism with wide-reaching consequences ever does occur, I pledge to go back and fix everything and we could restore the protection. But I highly doubt that it would be a problem. Your average vandal is not going to know about CFDW, much less attempt to cause mayhem through it, much less focus his efforts on disrupting the category system. Good Ol’factory ^(talk) 01:37, 18 August 2016 (UTC)[reply]
Unprotecting CFDW will need buy-in from the existing bot operators - as they are making edits based on these vetted entries from admins. Perhaps they would buy in to a CFDWLite page with less protection, but where they have a safety net in their bot jobs (e.g. no more than x edits per entry, no updates to protected pages) (would also require bot adjustments)? — xaosflux ^Talk 01:54, 18 August 2016 (UTC)[reply]
Personally, I wouldn't support this. Imagine if a vandal listed 1,000 categories and the bot did 10 edits for each one. Reverting would be a mess. Semi-protection or even 30/500 protection isn't enough for something like this given the potential for abuse. The non-admins running a bot themselves more tightly controls access and makes it much more difficult to vandalize on a massive scale. Moreover, unprotecting WP:CFD/W doesn't address the real issue here, which is that some editors don't even agree that non-admins should be closing CfDs as delete in the first place. We probably need to pose that question to the community and hope for the best. ~ Rob₁₃^Talk 02:06, 18 August 2016 (UTC)[reply]
Yeah, it's easy to imagine all sorts of chaos that could happen. I just happen to think it's extremely unlikely to ever occur. But fair enough, I can understand the hesitancy, especially from the existing bot operators. Good Ol’factory ^(talk) 05:26, 18 August 2016 (UTC)[reply]
@Good Olfactory: It's certainly a bit of a Catch 22 on this front. Since the lowering of the protection level was a revert of an admin action, WP:WHEEL would require strong consensus to lower the protection level again. Strong consensus requires a well-advertised discussion at a noticeboard, most likely. A well-advertised discussion greatly increases the risk of vandalism, so the protection level should not be lowered following such a discussion, even if it would be beneficial to lower it without such a discussion. ~ Rob₁₃^Talk 05:31, 18 August 2016 (UTC)[reply]
I can see that; good points. (Just ignore the last part of my musings above, which I have now put in small font. I can see the benefits of just leaving that issue aside.) Good Ol’factory ^(talk) 05:56, 18 August 2016 (UTC)[reply]
Security through obscurity is rarely a desirable practice. (For anyone considering this issue, the article is worth reading). --BrownHairedGirl (talk) • (contribs) 08:48, 18 August 2016 (UTC)[reply]
Very true. I'm not arguing for it, just arguing that if we can't even have obscurity, there's truly no chance. ~ Rob₁₃^Talk 08:50, 18 August 2016 (UTC)[reply]
I would more classify it as "security through apathy" than obscurity. But I guess the apathy is born of obscurity .... Good Ol’factory ^(talk) 22:56, 18 August 2016 (UTC)[reply]
Having recently finished manual reversion of a series of CfDs which were overturned at DRV and RfC, I recoil in horror at the thought of unprotecting CFDW. The work involved in reverting mistaken bot edits can be very burdensome. – Fayenatic London 08:21, 22 August 2016 (UTC)[reply]
Now I understand the background of this request, it was meant to be able to count how many non-admin closures were done ~~in 2015~~. From my side, in 2015 [added], the number may have been anywhere between 100 and a couple of hundreds, I don't remember and don't really bother about the exact number. In any case there was a huge backlog all the time and I thought I might be of help. Marcocapelle (talk) 18:18, 18 August 2016 (UTC)[reply]
Oppose this task, as I agree with BrownHairedGirl here. I'm all for non-admins helping out when they have the technical ability to do so, but this isn't the case here. If SSTflyer wants to be more useful closing CFD's, I would advise her to look here. -- Tavix ^(talk) 18:29, 18 August 2016 (UTC)[reply]
This whole discussion doesn't lead anywhere if there's no willingness to face the problem. The statement is being made that admins should close CfD discussions. Fair enough. Fact is too few admins do close CfD discussions. Then what? Marcocapelle (talk) 21:14, 18 August 2016 (UTC)[reply]

I've been lurking at CFD for a couple weeks now trying to learn the ropes, and I've slowly begun to work on the backlog. Yes, there are never enough admins, but in my eyes that should encourage people to run at RFA. As far as a "problem", there's backlogs, yes, but that's just part of Wikipedia. There's nothing really urgent in CFD that says a discussion needs to be closed after a week or two. Here's a good essay on the topic. Just relax, the problem will take care of itself. It always has. -- Tavix ^(talk) 21:53, 18 August 2016 (UTC)[reply]

Thanks I really appreciate you're willing to join! I wish more admins would follow your example so we wouldn't need to have this discussion. Just to answer your point, please note that I know from experience that problems do not take care of themselves when there is a structural lack of people. In 2015 Good Olfactory became inactive (in retrospect it was just for some four months, but nobody could know that in advance) resulting in the backlog growing and growing every day, while regular notifications at the administrators' noticeboard about the situation remained unanswered. During those months Fayenatic london was the only 'permanent' admin at CfD who did all he could to keep things a bit under control, but it was obviously more than he could handle - even with some help from my side as a non-admin. Marcocapelle (talk) 05:57, 19 August 2016 (UTC)[reply]

Comment. Having a non-admin with a bot that does CFDW-style work would not be unprecedented. There is User:ArmbrustBot, which can do such work, but only when activated by User:Armbrust, who is a non-admin. Good Ol’factory ^(talk) 22:59, 18 August 2016 (UTC)[reply]
Yeah, that's true. If anybody needs help, they just need to make a request at my talk page. Armbrust ^{The Homunculus} 22:41, 20 August 2016 (UTC)[reply]
Oppose, because Cat-a-lot does not have a facility to leave a link to the CfD discussion in the edit summary. I have used it myself for CfD implementation, particularly when merging to multiple targets, but ArmbrustBot is better for that purpose too. The option to ask Armbrust is a good option for both admins and non-admins. I recently also gave chapter and verse somewhere on 2012 precedents for a non-admin (me, at that time) closing CfDs to be implemented by an admin; I'm currently away from home and can't remember where I wrote that up, but I do vouch for what Rob13 wrote on this matter above. I do welcome and highly value the contributions made by non-admins in closing CfDs, and applaud SSTflyer's interest in doing this in a way that is as considerate as she can be of other editors. – Fayenatic London 08:47, 22 August 2016 (UTC)[reply]

The above discussion is preserved as an archive of the debate. Please do not modify it. To request review of this BRFA, please start a new section at WT:BRFA.