Jump to content

Wikipedia:Arbitration/Requests/Motions

Edit links
Page semi-protected
From Wikipedia, the free encyclopedia
< Wikipedia:Arbitration‎ | Requests

This is an old revision of this page, as edited by GoldenRing (talk | contribs) at 13:44, 5 July 2018 (Motion enacted). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

Arbitration Committee proceedings

Currently, there are no requests for arbitration.

[edit]
Open cases

Currently, no arbitration cases are open.

[edit]
Recently closed cases (Past cases)
Case name Closed
Venezuelan politics 25 May 2024
Request name Motions  Case Posted
Amendment request: Article titles and capitalisation none (orig. case) 7 June 2024
Clarification request: mentioning the name of off-wiki threads none none 4 June 2024
Amendment request: India-Pakistan none (orig. case) 7 June 2024
Clarification request: Contentious topics restrictions none none 10 June 2024
Motion name Date posted
Motion: Discretionary Sanctions 5 July 2018

Motions

Motion: Discretionary Sanctions

The following sentence is added to the end of the "Alerts" section of Wikipedia:Arbitration Committee/Discretionary sanctions: "Editors may not use automated tools or bot accounts to issue alerts."

The Arbitration Committee is aware of a discussion taking place at the Village Pump regarding issuing discretionary sanctions alerts via bot. As this discussion has a potentially large impact on how discretionary sanctions operate, the Arbitration Committee has decided to clarify existing procedures to note that alerts are expected to be manually given at this time. This is intended as a clarification of existing practices and expectations, not a change in current practice. The Arbitration Committee will fully review the advisory Village Pump discussion after completion and take community comments under consideration.

Enacted - GoldenRing (talk) 13:43, 5 July 2018 (UTC)[reply]

For this motion there are 12 active arbitrators, not counting 1 recused. With 0 arbitrators abstaining, 7 support or oppose votes are a majority.

Support
  1. Thanks to BU Rob13 for the text of this motion. Delivering DS notifications by bot has many logistical hurdles including but not limited to biting newbies, not being able to include every single page needed, and the personal reply to these alerts that can occur. -- Amanda (aka DQ) 19:24, 3 July 2018 (UTC)[reply]
  2. ~ Rob13Talk 19:32, 3 July 2018 (UTC)[reply]
  3. ♠PMC♠ (talk) 19:26, 3 July 2018 (UTC)[reply]
  4. Doug Weller talk 19:31, 3 July 2018 (UTC)[reply]
  5. Per my comments below. Newyorkbrad (talk) 19:41, 3 July 2018 (UTC)[reply]
  6. Katietalk 20:00, 3 July 2018 (UTC)[reply]
  7. RickinBaltimore (talk) 21:43, 3 July 2018 (UTC)[reply]
  8. Comments below. Mkdw talk 19:28, 4 July 2018 (UTC)[reply]
Oppose
Abstain
  1. Alex Shih (talk) 20:00, 3 July 2018 (UTC)[reply]
  2. It's already passed and I think I might be a crazy person, so I'm not going to try to swim upstream. I don't think this is all that clear - there's always doubts about what exactly is "automated" versus permissibly only "semi-automated" - but I also don't think this is all that important to add to an already-bloated set of rules. I know there have been some concerns about "biting", but speaking for myself, if I am in the early stages of joining a new community and am still feeling things out, I would much rather get an automated notification than have someone specifically make the effort to come tell me about all the things I'm probably going to screw up. I know the response "but then you have someone to ask!" but if I'm new and just learning I'd much rather get feedback from a bot than have someone personally involved in my screwups. I realize I seem to be in the minority on that point, but since my baseline preference is very much for automated notifications and against the "but it's more personal!" argument, I think it's fair to put that on the record. (All that being said, in all the vast amounts of text I've seen on this subject, I've never seen anyone really put forth a convincing argument that automated notifications that aren't too big of a nuisance are actually doable. Take one topic area, write the prototype, put the output on a sandbox page indicating who would've gotten notifications and what prompted them, and then when there's a working proof-of-concept we can talk.) Opabinia regalis (talk) 07:29, 5 July 2018 (UTC)[reply]
Comments by arbitrators
  • I want to particularly emphasize that we will review the RfC in its entirety and take all comments on board. ~ Rob13Talk 19:31, 3 July 2018 (UTC)[reply]
  • Comments:
I support the motion, but I think it is important to remember how we got to this point.
The concept of "discretionary sanctions" (DS) grew out of what was once called "article probation." It recognizes that certain topic-areas are unusually controversial or suffer from a high level of editing disputes, which our usual dispute-resolution mechanisms are not resolving in a timely way. ArbCom therefore gives administrators greater discretion than usual to address problems in these areas, such as by disallowing specific editors from participating in those areas, or by modifying the ordinary policies and guidelines with respect to them (e.g. 1RR instead of 3RR).
The first time I voted as an arbitrator to authorize DS in a topic-area, I pointed out that it was only fair to provide editors with warnings before imposing sanctions on them. As I've reiterated many times in the ten years since then, this is based on ordinary concepts of fairness and common sense. For example, an editor who makes two reverts on a page will be confused if she is reprimanded or blocked for breaching 1RR, if she had no idea that the page was subject to 1RR, and the like.
I certainly did not intend, when I first suggested "where possible, please warn before blocking or topic-banning," for ten years of accumulated rules-creep—the inevitable result of ten years of wikilawyering in our most controversial subject-areas—to yield a complicated set of "awareness criteria" and frequent arguments about who is or isn't "officially aware" that a given topic or page is subject to DS, or what all the rules are. Not for the first time, I suggest that all concerned read this article, by a current member of the WMF Board of Trustees, on how the increasingly bureaucratic and quasi-legalistic aspect of how Wikipedia operates may be deterring new editors and be unsustainable.
Administrators (or experienced non-administrators) should make editors aware that discretionary sanctions apply to a page or a topic-area when an editor appears to be editing problematically, or more generally, when the editor needs to be aware that his or her editing on a particular topic is subject to special rules. Good judgment should be observed in deciding when such a notification is warranted, and I agree with my colleagues that at this stage it is better that the notification be given by an editing colleague and not automatically by a bot. Newyorkbrad (talk) 19:57, 3 July 2018 (UTC)[reply]
This is a good clarification, but both the timing and the wording may be construed as discouraging the RfC, which really isn't the case. The RfC proposal isn't really concrete at the moment, so there's not much to comment about over there; but if we ever do automated notifications for DS, it would probably be done through centralised account; it makes no sense for individual editors to be issuing alerts through automated tools, so it would be another reason to justify this clarification. Alex Shih (talk) 20:00, 3 July 2018 (UTC)[reply]
User:Pbsouthwood, we had a discussion about edit boxes earlier this year and people pointed out that editors on mobile devices will not see the edit box notice. Wikipedia:Arbitration Committee/Procedures#Page restrictions was then amended to add "Editors using mobile devices may not see edit notices. Administrators should consider whether an editor was aware of the page restriction before sanctioning them." Doug Weller talk 12:54, 4 July 2018 (UTC)[reply]
  • I have read over the VP discussion and it is nearly evenly split between both sides. It is apparent to me that further discussion and exploration should take place to continuously improve the efficiency and stability of how discretionary sanctions work. It is an imperfect solution to an even more complex problem. I am supporting this motion as a point of clarity that discretionary sanction alerts are to be delivered by the direct and specific intent of another fellow editor. I have no intention of targeting or taking away tools and scripts that editors use to assist them when conducting these tasks, but fully-automated tools and bots are not appropriate here. I have serious concerns about having these alerts delivered en masse to editors without a guiding human hand behind each alert delivered. The implications greatly exceed the protections offered by disruptive alerting policies and would fail to address the principle issue here, that Wikipedia relies on collaboration and judgement by administrators and experienced editors when implementing and enforcing discretionary sanctions. Alerts are an important part of that process.
I would like to thank Newyorkbrad, NewsAndEventsGuy, and Boing! said Zebedee who I felt delivered some of the best comments and arguments in the discussion. Mkdw talk 19:28, 4 July 2018 (UTC)[reply]

Community comments

  • I do indeed hope that the Committee will carefully consider community input from the VP discussion. In particular, I want to draw to your attention one of the primary reasons for interest in such a bot: the very wide perception among editors receiving an alert that the language about not implying any wrongdoing is just boilerplate and should be disregarded, and that the real reason for the alert is that it is threat to start an AE complaint. Many times, I and other editors have made suggestions at ArbCom talk pages about revising the content of the alert templates, and each time, the discussion passes into archivehood without any explanation of why the Committee has not done anything. Maybe you are too close to one "side" of this process to see it, but it really is a significant problem that needs addressing. (Look at it this way: if you value the personal touch of an editor deliberately issuing the alert, then you should be interested in the feedback of those editors.) If not a bot, then how the alerts are formulated. --Tryptofish (talk) 19:44, 3 July 2018 (UTC)[reply]
    • For me, it’s stalled because nothing materially better has been proposed. I’ve seen a million copy edits that rearrange some of the words, but never while changing the tone or meaning. The authors usually think they’ve fundamentally altered the text, mostly because they ascribe the tone they were going for to their edit, but I’ve yet to be convinced anyone looking in from the outside would notice any difference. ~ Rob13Talk 19:50, 3 July 2018 (UTC)[reply]
  • @Brad: I read the article you linked to. And I can see the logic leading to preferring manual notification rather than a bot. But I think a further implication of wanting to make Wikipedia more inviting to new editors would be to make the alerts less nasty looking and more informative. That would not be adding more bureaucracy, but rather making the process easier to understand correctly. --Tryptofish (talk) 20:04, 3 July 2018 (UTC)[reply]
  • Good motion, the ideal outcome. Pinging Bellezzasolo as this would seem to prohibit their DS tool. ~ Amory (u • t • c) 20:36, 3 July 2018 (UTC)[reply]
    • (edit conflict) Hm, that tool simplifies the process by checking for previous alerts, but doesn't truly automate giving a DS alert (ie you can't set it to "alert everyone who edited in American Politics today"), so in my opinion it should be allowed. We might need to adjust the wording. BU Rob13, I know you're on a plane, but any thoughts? ♠PMC♠ (talk) 20:42, 3 July 2018 (UTC)[reply]
      • Just landed. That’s an example of a semi-automated tool, a tool where the script does the heavy lifting but an editor approves each edit. It is not automated (used on Wikipedia to mean “edits done by a program or script without direct editor supervision”) and isn’t covered by the motion. ~ Rob13Talk 20:51, 3 July 2018 (UTC)[reply]
  • (edit conflict) BU Rob13, does this cover twinkle extensions? I use User:Bellezzasolo/Scripts/arb.js, which still forces the edit filter warning, but is easier than having to look up the codes (it also generally can tell if someone has been alerted before.) Otherwise, I oppose the bot, so obviously I’m in favour of clarifying this for now, which would have the effect of any bot needing permission by the committee to operate. TonyBallioni (talk) 20:38, 3 July 2018 (UTC)[reply]
    • It seems to me that the proposed language of "automated tools or bot accounts" should probably have "automated tools" removed or clarified. The distinction between "automated" and "semi-automated" strikes me as not useful. --Tryptofish (talk) 21:52, 3 July 2018 (UTC)[reply]
      Especially since we don't actually make such a distinction in many places; it's not consistently recognized in guidelines, policies, ANI decisions, yadda yadda.  — SMcCandlish ☏ ¢ 😼  23:36, 3 July 2018 (UTC)[reply]
  • It's disappointing to see the burden of Arbcom's rickety alert process shifted onto the shoulders of editors, simply because a new editor might feel bitten. So, reword the alert—problem solved. The alert rules are unduly bureaucratic, confusing, contradictory, and hard to locate. I would like for Arbcom to reconcile these two seemingly contradictory rules from Wikipedia:Arbitration_Committee/Discretionary_sanctions#Awareness_and_alerts
    1. No editor may be sanctioned unless they are aware that discretionary sanctions are in force for the area of conflict. An editor is aware if In the last twelve months, the editor has given and/or received an alert for the area of conflict.
    2. Editors issuing alerts are expected to ensure that no editor receives more than one alert per area of conflict per year. Any editor who issues alerts disruptively may be sanctioned.
As someone who has alerted quite a few editors, I am telling you that this process is fraught with problems and it's onerous. A bot would a step forward. Alternatively, revising the alert rules so that a notice on an article talk page is sufficient would address most of the concerns.- MrX 🖋 21:08, 3 July 2018 (UTC)[reply]
  • Those two statements aren’t contradictory. After the year is up, you alert the editor again. I wouldn’t mind doing away with the “in the last twelve months” bit entirely, so long as we also indicate we expect enforcing admins to consider whether the editor was genuinely not aware of DS even though they technically received an alert ages ago. ~ Rob13Talk 21:22, 3 July 2018 (UTC)[reply]
  • For crying out loud, alert them once and be done with it!. All this alert renewal nonsense does is enable wikilawyers to game the system and prevents the average editor from availing themselves of the arbitration enforcement process. - MrX 🖋 01:42, 4 July 2018 (UTC)[reply]
  • Absolutely. And the whole "awareness" thing is silly in the first place, since the affected pages already have (or, should have and really immediately need to have) the appropriate editnotices and talk page notices (and talk page editnotices), though perhaps less scary ones. No one can keep editing the same stuff seeing notices about DS and not be aware of DS. The user-talk-delivered alerts are superfluous. I don't buy the BITE arguments against the Ds/alert bot for a second, because the editnotices are an order of magnitude more menacing a present than the user-talk ones. I mean, seriously. Just go load Donald Trump in edit mode. It looks like you're in for a visit from an FBI SWAT team for daring to think about editing the page. I don't want to hear any patent nonsense about there being a marked difference between bot or hand delivery of the exact same shorter, blander Ds/alert. It's like trying to decide which whisker on your cat is longer, while your house is on fire. this is a ranging fire of "New editor: Run away as fast as possible."  — SMcCandlish ☏ ¢ 😼  05:04, 4 July 2018 (UTC)[reply]
It is quite common to edit an article without reading the talk page first. It is also quite difficult to prove that someone has read a talk page, so a warning on the talk page is worth very little as verifiable notification, unless the person edited the talk page too. A banner that displays in the edit box will be seen by everyone who edits, unless they are blind, in which case their text reader will read it out to them. Anyone intending to enforce DS can quite reasonably assume that in such a case, the person was notified, and as recently as is possible.· · · Peter (Southwood) (talk): 11:37, 4 July 2018 (UTC)[reply]
Doug Weller, Thanks for clarifying that important detail (that mobile users do not see the edit box notices) - I do not use mobile, so had not noticed. I guess that this is a thing that could be fixed by the developers if it was considered necessary or sufficiently useful. · · · Peter (Southwood) (talk): 13:18, 4 July 2018 (UTC)[reply]
Might even be fixable without them; depends on why mobile users aren't seeing it. Is it because the server is ignoring everything in editnotices for the mobile version, or is it because of how the edit notices are coded?  — SMcCandlish ☏ ¢ 😼  19:50, 4 July 2018 (UTC)[reply]
  • I don't really see the point of this motion. It's wikilawyerly nitpicking. No one suggested that at present it's okay for a bot to deliver this notice. The entire point of the WP:VPPRO proposal is to change that (or to have the community come to a consensus on some other-direction improvement to the current DS system). I mean, really. Given everything of substance actually being said in that RfC, the Arbs' only response is to say something tautological about the present? That's a really bad sign.  — SMcCandlish ☏ ¢ 😼  22:29, 3 July 2018 (UTC)[reply]
    • I share that concern. It feels like ArbCom felt an urgent need to, in effect, shut down the VP discussion, as though things were about to get out of hand and there were an urgent need for getting it under control. I think that's part of what Alex Shih said above. There's no emergency here, and the Committee might want to carefully examine the wording of the motion before enacting it. --Tryptofish (talk) 22:48, 3 July 2018 (UTC)[reply]
  • On this specific motion and its interpretation: Suppose I decide too many people are being asshats in articles covered by WP:ARBAP2 [which I think we all agree is true]. I use some analysis tools to ID everyone who's edited ARBAP2 articles or their talk pages more than X times in Y timespan, and then use AWB to deliver them all a {{Ds/alert|ap}} (which for each requires a manual save, and will pop up the are-you-sure editfilter box with the steps for checking that they don't already have one this year). Permissible?  — SMcCandlish ☏ ¢ 😼  00:47, 4 July 2018 (UTC)[reply]
    • It doesn’t conflict with this particular clause but may be sanctionable under the sentence that prohibits disruptively delivering alerts if the community or the Arbitration Committee feels that editors are being bitten. ~ Rob13Talk 01:54, 4 July 2018 (UTC)[reply]
      • That's just subjectively arbitrary (in the bad sense), though. Either it's okay to use a tool that requires manual examination and saving (i.e., no one even knows you're using the tool, unless they look through filters to find out), after you've complied with the instructions and made sure you're not leaving a duplicate notice within the year); or it's not. If it's not, on what basis? Who cares, and why? It would be easy enough to do exactly the same thing by manually going down the list. If the template is, in your view, intrinsically BITEy for new editors (the only ones to whom BITE pertains), then there's deeper problem here than using this tool or that or using no tool. I don't have any plans to do this, with or without a tool, of course. I'm just trying to get a straight answer. If there are rules surrounding these templates, and potential sanctions for not following them, they have to be clear, and they have to more clear the more we automate and semi-automate over time. That's a one-way trip; no one is using fewer tools and doing things the slower, harder, more human-error prone way after they've already discovered assisted ways to get the jobs done better.  — SMcCandlish ☏ ¢ 😼  04:57, 4 July 2018 (UTC)[reply]
        • There will always be some judgement calls. If you notified every editor on the site about DS in a topic area with AWB, you’d be sanctioned. That’s because you notified all editors on the site, though, not because you used AWB. That’s all I’m saying. I’m not “coming for your tools”, and I fully agree with Mkdw on the topic of tools/scripts/etc. I would frown on large-scale indiscriminate notices, though, as they lose the human element (and human determination of cost/benefit to notifying). (WP:MEATBOT would also be relevant in the extreme case I mention above.) ~ Rob13Talk 20:48, 4 July 2018 (UTC)[reply]
  • I believe that discretionary sanction notices should be given to users manually by other users and by using the templates provided in the procedure page, or with the guided assistance of a tool (like Twinkle) - one that requires the user to choose from a drop-down list of active discretionary sanctions, the article that the recipient is/was editing to warrant the notice, and guardrails to prevent and disallow common mistakes be made - so that an accurate and proper notice is correctly and appropriately left the first time, and every time.
No bots should be allowed to leave discretionary sanction notices, and no "one click and drop" (AKA "drive by shooting") automated processes should be allowed to be developed to carelessly shortcut this very sensitive, delicate, and serious procedure. The notice in itself is quite scary-looking, overwhelming to read, and (in my experience) will frequently startle, scare, and chase away new users who are simply not aware of what "ArbCom", "enforcement", or "discretionary sanctions" even are. To allow for automated tools to take place and simply allow for someone to plop an AC/DS warning without careful checks and requiring the user to fill in the specific options necessary would be a reckless decision to allow... and to allow the development of bots to do this automatically will result in this procedure becoming a careless, thoughtless, and routine event where editors "click and go" without giving a second thought, and bots "plop and drop" without any regard to the situation on a case-by-case bases. My experience as well as what's clearly stated on multiple policies and guideline pages within ArbCom - clearly show that this procedure is a very serious matter where very careful thoughts and consideration must be taken into account every time one is left... and allowing for careless automation and for bots to take this place would lead to this procedure being treated the opposite. ~Oshwah~(talk) (contribs) 03:33, 4 July 2018 (UTC)[reply]
Why should you need to use a GUI drop-down if you already know the correct code? Not everyone favors mouse-based tools (and for some, they are an accessibility problem). If we all took this "Ds/alert is very very serious, and requires dreadful amounts of consideration" stuff to heart (which ArbCom never has to date, so far as I've seen), then they need to be redesigned and redocumented. We can't have it both ways. Either Ds/alert is a no-wrongdoing-implied notice that different rules apply to a topic, or it is a warning of potential sanctions just like a level-3 or higher WP:UWT template is.  — SMcCandlish ☏ ¢ 😼  05:14, 4 July 2018 (UTC)[reply]
SMcCandlish - I wasn't trying to state that a drop-down menu and the use of the mouse was the actual and only solution that we should accept; I was merely trying to give an example as to how an automated tool could or should be designed when this process is in mind. Above all, all I wanted to emphasize is that the tool shouldn't be written to allow a careless "drive by click-and-run" or a "one click plop-and-drop" involving AC/DS warning templates, and instead should be formulated to accept and sanitize user input and care. ~Oshwah~(talk) (contribs) 10:07, 4 July 2018 (UTC)[reply]
  • I've commented at the RFC, but my core objection to the suggestion that bots should deliver DS alerts to everyone editing related pages is simply that those editing well should not receive those alerts at all. Any alert, however well written, has the potential to be discouraging (especially to an inexperienced editor), and those who are not causing problems don't need to be told anything about DS. DS alerts should be only for those getting close to causing problems, which I believe was the original intention. Boing! said Zebedee (talk) 04:45, 4 July 2018 (UTC)[reply]
    If this were true (or true enough that we care – do we care about, say, a 1% discouragement if it stops 5% of impending disruption?), then the editnotices at the articles would be where the concern is. This is another of the "can't have it both ways" things. If DS-related notices are inherently scary and discouraging, we're scaring off way more editors every day with the editnotices than we ever have with Ds/alerts the entire time Ds/alert has existed. Editors are hit in the face with the editnotices every single time they edit any of those pages.  — SMcCandlish ☏ ¢ 😼  05:18, 4 July 2018 (UTC)[reply]
    I don't think you should be bandying about 1% or 5% figures when you don't actually know them - and how do you know that "we're scaring off way more editors every day with the editnotices than we ever have with Ds/alerts the entire time Ds/alert has existed"? And any valid comparison would be between the existing system and your bot-delivered notices to everyone - not between current edit notices and current manual DS alerts. You know there's a difference between a personal alert on a user talk page, a general alert on an article talk page, and an edit notice, and comparing them needs to be treated very cautiously at best. Finally, and I'm just being honest here so I hope you don't mind, your constant badgering is starting to make your obviously well-intentioned proposal look increasingly like a soapbox rant. That's all I'll say - please slow down the talking a bit, and listen to and think about other opinions too without going straight into argument mode every single time. Boing! said Zebedee (talk) 09:04, 4 July 2018 (UTC)[reply]
    I'm not "bandying about" anything, I'm using obviously hypothetical numbers to suggest an analysis needs to be done to find out what the numbers really are and what they mean. Or, in the absence of analysis, that assertions that editors are being driven away by Ds/alert are baseless supposition. Especially since they get way harsher DS-related "scareboxes" far more frequently from another direction. I also don't need to have my posts here or at VPPOL mischaracterized by you. I'm not just telling people they're wrong, I'm asking questions, pointing out where the proposal has already answered theirs, agreeing with their concerns and clarifying the RfC wording to address them (still ongoing), and many other things. I'm managing the proposal, not WP:BLUDGEONing it.  — SMcCandlish ☏ ¢ 😼  19:58, 4 July 2018 (UTC)[reply]
  • Posting here to solicit opinions from editors/arbitrators who have contributed to the relevant discussions. I am intending to remove one header from {{Ds/alert}}, see Template_talk:Ds#Template:Ds/alert. Would like to request some input to see if it's a good start to make the alert template looking more like a notice as intended, instead of a warning that alleges wrongdoing. Alex Shih (talk) 08:28, 4 July 2018 (UTC)[reply]
  • Oppose over-reaching knee-jerk response "automated tools" is extremely vague, what does this mean? If I wanted to send 2 people a notice, and used AWB to do it, but in semiautomated mode is this a "automated tool"? What about something like a browser script? How about copy-and-pasting? I don't have any major objection against using a "bot account" in principal though. — xaosflux Talk 14:47, 4 July 2018 (UTC)[reply]
Wikipedia has a fairly documented language about the difference between automated and semi-automated tools. Nearly all the examples you've provided are examples of semi-automated editing and would not be restricted by this motion. Mkdw talk 17:55, 4 July 2018 (UTC)[reply]
And edits are the responsibility of whomever makes them. Why the arbitration committee is interfering with the "method" of the edit, rather than the content is beyond me. We already have provisions for dealing with disruptive alerting in general. Especially in the case of newer editors avoiding "bot account" edits may be useful as the new editors may discount it due to the non-"person" nature of the edit. — xaosflux Talk 18:08, 4 July 2018 (UTC)[reply]
Even if there is documented language about the difference between automated and semi-automated, it's still an ambiguous distinction. At a minimum, I strongly recommend blue-linking "automated" to something that defines the difference. --Tryptofish (talk) 19:30, 4 July 2018 (UTC)[reply]
For sure, a footnote or a link with further details would be appropriate. Mkdw talk 19:36, 4 July 2018 (UTC)[reply]
@BU Rob13: This has come up several times in this discussion, so something extra to clarify this point seems like a good idea. Mkdw talk 19:36, 4 July 2018 (UTC)[reply]
Yes. My questions above at the "On this specific motion and its interpretation" comment block remain unaddressed. BU Rob13's conceptualization of this and yours, Mkdw, don't appear entirely compatible.  — SMcCandlish ☏ ¢ 😼  20:05, 4 July 2018 (UTC)[reply]
The other issue is that this is making a rule that is hard to impossible to review violations of. If I write a utility that lets me target a username, then it "automatically" logs in as me, edits that user talk vs going to the webui and pasting in the ds alter template what is the issue? You can "manually" send a ton of these with tabbed browsing in short order with that exact same impact. — xaosflux Talk 20:13, 4 July 2018 (UTC)[reply]
@Mkdw: my whole point is that Editors may not use automated tools or bot accounts to issue alerts. is just making up a rule that isn't needed and can just be an avenue to future arguments. Isn't the point that the committee doesn't want these alerts coming from bots? e.g. "Editors may not use bot accounts to issue alerts." should suffice? — xaosflux Talk 20:24, 4 July 2018 (UTC)[reply]

Comments by SMcCandlish (the RfC proponent)

As some of you will be aware, I've tried to address various issues with DS and Ds/alert at ArbCom talk, ARBPOL talk, AC/DS talk, even Ds/alert talk, over many years, and from many angles, and been stonewalled on virtually every single point, every time. Even, back when, getting the Ds/alerts revised to stop making bogus accusations of wrongdoing in them (that shamefully took multiple ARCAs, one of which had to be opened on my and someone else's behalf because ArbCom – a compositionally different ArbCom – just wouldn't listen otherwise). So, this RfC was not some off-the-cuff whim. It's been years coming, and there really was no other option to crack through the crust of ArbCom's bureaucracy bubble; cf. second comment here. (I am of course aware that ArbCom changes membership somewhat from year to year; this is an institutional problem, not a personalities one. I ran for ArbCom last term, with the primary goal of addressing this stuff hands-on. I received more support than several now-sitting Arbs, but I have some detractors from various content disputes, so I didn't quite make the cut due to opposes.)

  • @Newyorkbrad: Thank you for the thoughtful comments. I think your summary leaves out some important things: DS was invented, out of thin air, by a single Arb. It was then just foisted off on the community, and we've never been all that comfortable with it, especially as its scope has crept multiple ways. It's also why WP:RFA turned from challenging into a hellhole: DS increased the trust level we have to have in an admin by an order of magnitude. There was a "review" of DS several years ago, but sculpted by another single Arb, who steered it very carefully to produce results that just reinforced that DS was here to stay, and to stay in pretty much exactly the same form with a few twiddly tweaks.

    DS has crept from encyclopedia content disputes into interfering with internal policy debate, a separation-of-powers problem. And it's gone from something to implement when really needed, for only as long as really needed, into something that never, ever goes away once authorized for that topic, as long as someone can show "a dispute" here or there, despite the fact that the disputes pointed to do not need DS to resolve them but could be handled at ANI, and despite attempts to use DS to do so turning into horrible messes that would not have happened at ANI (or NORN, ANEW, whatever). It's the classic non-fallacious slippery slope: once power is taken ("temporarily", "for an emergency") it will not be yielded. When a civil society collapses into despotism over time, this is almost always how it happens. Jemielniak's op-ed isn't spelling it out in the same terms, but these ideas are of a piece; coups happen more often by an privileged, insider bureaucracy working slowly than by a sudden violent upheaval.

    But let's suppose DS is just fine as-is. The current "awareness" cognitive dissonance stuff is not. I've delivered a lot of Ds/alerts. To date, only one single time has someone treated it as an awareness notice instead of an attack/threat to which they should respond with escalation or even public grandstanding in article talk about being "attacked". If we don't have a bot deliver them automatically, then we need a bot that does it by request. Or an admin board where admins go do it by request. Or a button we can click that does it by request. Without delivery being tied (at least without digging around) to someone specific. Otherwise, 99.5% (or so) of the time it's going to continue to escalate rather than reduce conflict. It's having the diametric opposite of the intended effect. Counterproductive. Failure. Whatever words you like for "not working".

    An even saner plan is to just put DS editnotices and talk-page banners on all affected pages, and treat these as "awareness". If you edit there, you get the notice, you are aware, the end. Or have some micro-bureaucratic rule, if you must, like there must be X edits to such pages in Y span of time to be aware; that could be reasonable. Or – this isn't crazy – forget the awareness stuff and just enforce the rules evenly, like we do with all other rules (and with leeway shown to new editors). None of this is innately hard. ArbCom has forced it to be hard against the community's expectations and collective will.
     — SMcCandlish ☏ ¢ 😼  23:18, 3 July 2018 (UTC)[reply]

    PS: We "should make editors aware that discretionary sanctions apply to a page or a topic-area when ... the editor needs to be aware that his or her editing on a particular topic is subject to special rules." That's more tautology. Why would an editor need to have this fact hidden from them? Sounds like a great argument in support of my RfC proposal, frankly.  — SMcCandlish ☏ ¢ 😼  23:20, 3 July 2018 (UTC)[reply]

  • @Alex Shih: Thank you for weighing in as well. "The RfC proposal isn't really concrete at the moment" – Indeed. It's a first-draft approach (aside from minor textual twiddles). I expected this to to be 80% oppose right off the bat, but it actually has legs, probably for a second, refined RfC another time. If the literal idea of having a bot drop off the notices ... well, robotically is flat-out rejected, it's very clear that the community thinks some of problems inspiring the proposal need to be addressed. I don't really care how that happens. The bot idea is more of a kludge than anything, a "better than nothing" patch, but if talking about it gets a productive ball rolling, great.  — SMcCandlish ☏ ¢ 😼  23:27, 3 July 2018 (UTC)[reply]
  • @Mkdw: Also appreciate your thoughtful comments on this, especially "I have no intention of targeting or taking away tools and scripts that editors use to assist them when conducting these tasks". It's not clear if this squares with BU Rob13's view on it, which seems tied to WP:BITE and perhaps to an over-extension of BITE to non-new editors, though it's hard to tell from this. I've asked for some clarifications [1].  — SMcCandlish ☏ ¢ 😼  20:09, 4 July 2018 (UTC)[reply]
  • @Doug Weller: Re, "Editors using mobile devices may not see edit notices. Administrators should consider whether an editor was aware of the page restriction before sanctioning them." – How does this square with the idea that the only recognized awareness comes from Ds/alert (or from direct involvement in "process" that makes the awareness certain)? If ArbCom and WP:AC/DS are maintaining the position that awareness only happens per WP:AC/DS#Awareness, then whether an editor had seen an editnotice or not would be just be irrelevant. It seems weird – rules-interpretation conflicted – for ArbCom to tell admins to "consider whether" editnotices constitute awareness for a particular user, while at another page disavowing them as awareness, ever, in the first place. I agree that the mobile problem right now makes it technically impractical to treat the editnotices as "awareness delivery" (something I and many others have proposed, in lieu of just dropping this "awareness" stuff); but for all we know, that might be something that could be fixed this very month. So, that's a separate matter from what really can constitute awareness at the moment.  — SMcCandlish ☏ ¢ 😼  20:53, 4 July 2018 (UTC)[reply]
    • We require edit notices for page-level restrictions in addition to the alert. That’s what that sentence is referring to. ~ Rob13Talk 23:58, 4 July 2018 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Arbitration/Requests/Motions&oldid=848955466"