Wiktionary talk:Votes/2014-08/Debotting MewBot

From Wiktionary, the free dictionary
Jump to navigation Jump to search

Rationale[edit]

The bot repeatedly performed edits that lack consensus and are often not discussed anywhere before they start (see #Poorly discussed bot tasks). The bot blatantly violates WT:BOT, but I don't think many people really care. The part of WT:BOT violated is this:

I will ask around for consensus, perhaps at the Beer parlour, or
I will make sure that the task is so innocuous that no one could possibly object.

This would not be so bad if the bot edits almost always turned to be largely supported after they start, but that is not the case. One of the last such bot runs is replacing the call to "pedialite" template in the mainspace with "projectlink|wikipedia", exemplified in diff.

The owner of the bot is User:CodeCat. They know that there is some opposition to their method of procedure with the bot; this vote is not the first measure taken to deal with the problem (see #Measures taken) but rather a last desperate attempt after other measures, especially repeated discussions, failed.

CodeCat seems to think that when they have a taks for the bot in mind that is not proposed anywhere, then because it is unopposed at this time, it is thereby supported by consensus. Therefore, they seem to think that every single run of their bot is supported by consensus at the point at which it starts, since at that point no one has yet disputed the bot run. I could be wrong about this, but it seems to follow from Wiktionary:Beer parlour/2014/August#CodeCat pushing original_research. --Dan Polansky (talk) 09:52, 14 August 2014 (UTC)[reply]

Updated with references to other sections of this talk page. --Dan Polansky (talk) 20:21, 17 August 2014 (UTC)[reply]

Severity of the measure[edit]

Debotting MewBot is a measure designed to be of moderate severity. In particular, it does not involve desysopping CodeCat or removing their AWB rights. It is not meant as a way to punish; it is meant to prevent continuation of violation of the bot policy and continuation of undiscussed bot runs with significant opposition. Since AWB rights removal is not proposed, CodeCat will have the technical means to continue non-consensual mass editing even if the vote passes, which is a big if anyway. If the vote passes (again, a big if), it will be a signal that the community actually supports WT:BOT and that the silent implied consensus for the abandonment of open consensus-forming processes was not really there. --Dan Polansky (talk) 05:05, 15 August 2014 (UTC)[reply]

Measures taken[edit]

In #Rationale, I claim measures were taken to communicate to CodeCat that this way of procedure with the bot is not okay. What follows is the supporting evidence for my claim. It is probably incomplete, since to find anything in these horrible LiquidThreads on CodeCat talk page is a nightmare.

--Dan Polansky (talk) 06:22, 15 August 2014 (UTC)[reply]

Poorly discussed bot tasks[edit]

What follows are some bot tasks run by MewBot that, as far as I know, were either discussed in wrong places (Grease Pit for a change massively impacting mainspace) or not discussed at all. It serves as supporting evidence for #Rationale.

Caveat: This report is based on how I remember things and on quick on-wiki research. The bot page and bot talk page do not seem to contain anything like links to discussions. Finding out what was discussed and where is rather hard. Please, post refuting information below, such as links to discussions where the things actually were discussed.

Some Tasks Run by MewBot
Bot task Example Diff Series of Edits Diff date
Change genders like f|p to f-p Diff Series 8 May 2013
Add {{context| to context labels Diff Series 4 June 2013
Add lang=en to context templates Diff Series 16 June 2013
Change/reorder parameters on en-noun Diff Series 25 August 2013
Change {{form of|...}} to {{nominative plural of...}} etc. Diff Series 6 January 2014
Change {{mk-noun|g=m|…}} to {{mk-noun|m}} Diff Series 17 June 2014
Change {{pedialite}} to {{projectlink|wikipedia}} Diff Series 12 August 2014

Note that the bot flag was originally requested in Wiktionary:Votes/bt-2010-03/User:MewBot for bot status for adding Dutch inflected forms.

--Dan Polansky (talk) 20:19, 17 August 2014 (UTC)[reply]

Refusal to undo changes[edit]

CodeCat's typical modus operandi is as follows:

  • 1. They peform changes for which it is unknown whether they have consensus.
  • 2. If significant opposition emerges, they do not undo the changes, claiming that there is no consensus for undoing the changes.
  • 3. By doing so, they are capable of enforcing changes that have sub-majority support: if, say, 55% of people oppose the changes made by CodeCat, there would still, according to our tradition, be no consensus for reverting the changes. Thus, changes supported by 45% of people become fait accompli.

A case in point is change from {{pedialite}} to {{projectlink|wikipedia}}. They run it without a prior discussion. I challenged the run in Wiktionary:Beer parlour/2014/August#Not renaming template pedialite, using my time and attention, and also that of other editors. When the outcome of the discussion was that the changes should not have been done, I asked CodeCat to undo the changes in Undoing pedialite changes in mainspace on their talk page; they refused, claiming that "There is no consensus for replacing {{projectlinks|wikipedia|...}} with {{pedialite|...}}.". --Dan Polansky (talk) 20:38, 17 August 2014 (UTC)[reply]

Dan is misrepresenting the facts here. I never refused to undo the changes. He wants me to restore the older status quo, but the change he asks me to perform would not mean restoring the status quo before I made any changes, but rather an entirely new situation for which no discussion has been made beforehand. Since this is the very thing that Dan seems to be hammering at me about, I feel at a loss what to do here. I'm the evil destroyer of Wiktionary no matter what I do. And then Dan wonders why I think he's after me personally? —CodeCat 20:44, 17 August 2014 (UTC)[reply]
If I misunderstood, I apologize. I ask that you, at the very least, restore the state before the bot run, as for projectlinks vs. pedialite. By consulting the bot edit history, you should be able to do that, by identifying pages that the bot updated, and performing the reverse replacement there. If you really believe that the bot change cannot be easily undone, and if you also believed that before you started the run, that is even worse. --Dan Polansky (talk) 20:51, 17 August 2014 (UTC)[reply]
That is going to be very hard to do. Firstly there is no easy way for the bot to know which pages it has edited in that fashion, as I don't know of any way that bots can look at the edit history of a particular user. Secondly it cannot tell what the original state of the entry was, whether it used {{pedialite}}, {{pedia}} or {{projectlinks}}. On the other hand, I don't understand why you're asking me to undo this. Was Wiktionary better off when there were somewhat more transclusions of {{pedialite}} and {{pedia}}, and somewhat less of {{projectlinks}}. It seems like you're asking this more to make a point, and to give me yet another opportunity to trip up and fail to meet your expectations, just so you can point your finger at me once more and say "see, it's evil, we must kill it!". A self-fulfilling prophecy very much akin to a witch hunt. I can assure you that you can't just weigh my actions with a duck. —CodeCat 20:56, 17 August 2014 (UTC)[reply]
Is it accurate to say that you run the bot knowing that even if a very strong opposition arises, there will be no undo? --Dan Polansky (talk) 21:01, 17 August 2014 (UTC)[reply]
Yes, that is accurate. It's also accurate for very many edits made by other editors in the years past, whether by bot or by hand. But of course, if I do it, it's suddenly supposed to be scrutinised. I just count myself relieved that Wiktionary is run by consensus rather than Dan McCarthy here. —CodeCat 21:04, 17 August 2014 (UTC)[reply]
Your words just seem to confirm my description of your modus operandi; you don't seem to be disputing that description. Also, edits made to single pages by hand are easy to undo; you seem to be confusing things together. --Dan Polansky (talk) 21:07, 17 August 2014 (UTC)[reply]
Of course my words just confirm that. Your judgement of my guilt has already been made, and all you look for now is evidence to prove yourself right in the eyes of the masses. Nothing but a show trial. Again, I count myself lucky that you are not my judge. —CodeCat 21:09, 17 August 2014 (UTC)[reply]
However, no matter the outcome of this particular case, I think the general description I presented for your modus operandi is accurate. As far as I know, there is not a single case in which you would have undone the changes made by your bot; please present refutation if you have it. --Dan Polansky (talk) 20:58, 17 August 2014 (UTC)[reply]
Let me counter that by asking, please provide evidence that you would have undone any of the edits you have made in the past. I'm pretty sure that's unprovable. —CodeCat 21:00, 17 August 2014 (UTC)[reply]
I don't remember anyone asking me to undo my edits. I don't remember anyone complaining about my AWB runs. My AWB runs are documented in the archives of my talk page.--Dan Polansky (talk) 21:05, 17 August 2014 (UTC)[reply]
That's not what you asked. You asked if I would have undone the changes, not whether I did. There is no way for me to prove a possibility that did not happen. —CodeCat 21:06, 17 August 2014 (UTC)[reply]
Do you agree that you have never undone any changes made by your bot? --Dan Polansky (talk) 21:08, 17 August 2014 (UTC)[reply]
No. —CodeCat 21:09, 17 August 2014 (UTC)[reply]
So do you say that you have undone the bot changes at least once in response to strong opposition? --Dan Polansky (talk) 21:11, 17 August 2014 (UTC)[reply]
Yes, your "honour". —CodeCat 21:12, 17 August 2014 (UTC)[reply]
──────────────────────────────────────────────────────────────────────────────────────────────────── And could I see the links, or are they too hard to find? --Dan Polansky (talk) 21:13, 17 August 2014 (UTC)[reply]
Of course you can see the links. You should have no more trouble finding them than I would, so I wish you luck. —CodeCat 21:14, 17 August 2014 (UTC)[reply]
Am I correct that you will not be directing me to the case that you above confirms exists? Thus, would you agree that the existence claim you are making is left unproven? --Dan Polansky (talk) 21:18, 17 August 2014 (UTC)[reply]
Later: I have undone the undiscussed and opposed changes, restoring the use of {{pedialite}}; see User talk:DPMaid#Restoring pedialite. It was quite easy, but still laborious a bit, since running AWB is laborious. The claims made above by CodeCat about how hard it would be ("That is going to be very hard to do") show lack of sincerety and a bucketful of excuses ("I don't understand why you're asking me to undo this"). --Dan Polansky (talk) 09:32, 23 August 2014 (UTC)[reply]

vote[edit]

@Dan Polansky: are you going to open this vote? DTLHS (talk) 19:37, 22 August 2014 (UTC)[reply]

He doesn't have to open it. Once the start date arrives, anyone can "open" it by removing the yellow box and voting. --WikiTiki89 19:45, 22 August 2014 (UTC)[reply]
Correct. --Dan Polansky (talk) 22:12, 22 August 2014 (UTC)[reply]

Incident[edit]

Wiktionary:Requests for deletion/Others#Template:inflected form of did not show consensus for deletion. That did not prevent MewBot from replacing all uses of the template with other templates, as is currently apparent from https://en.wiktionary.org/w/index.php?title=Special:Contributions/MewBot&offset=&limit=5000&target=MewBot. --Dan Polansky (talk) 23:34, 3 August 2015 (UTC)[reply]

Incident - inh[edit]

Replacing {{etyl}} with {{inh}}, apparently with no discussion showing consensus (I could be wrong). Example edit: diff. --Dan Polansky (talk) 11:26, 27 September 2015 (UTC)[reply]

Incident - topics[edit]

MewBot went on to replace {{C}} with {{topics}} in the mainspace without first ascertaining consensus, e.g. in diff. In Wiktionary:Votes/2017-05/Templatizing topical categories in the mainspace, it subsequently became apparent that such consensus is indeed seriously lacking. --Dan Polansky (talk) 15:22, 5 July 2017 (UTC)[reply]