Wikipedia:Bots/Requests for approval/Yobot 24


 * 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 Symbol delete vote.svg Denied

Yobot 24
Operator:

Time filed: 21:43, Monday, June 1, 2015 (UTC)

Automatic, Supervised, or Manual: Automatic

Programming language(s): AWB

Source code available: open source

Function overview: Remove Persondata

Links to relevant discussions (where appropriate): Village_pump_(proposals)/Archive_122 concluded that "Consensus is to deprecate and remove."

Edit period(s): One time run

Estimated number of pages affected: 1.5 million pages

Exclusion compliant (Yes/No):

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

Function details: Straightforward

Discussion
Two questions: Thanks! GoingBatty (talk) 23:30, 1 June 2015 (UTC)
 * 1) Should this bot wait until AWB has been changed to stop adding/updating Persondata?
 * 2) Since Persondata is not visible in the article, does WP:COSMETICBOT apply?  Would it be better to include Persondata removal in AWB general fixes, for other bots & users to remove as they make substantial changes?


 * GoingBatty I'll be doing general fixes at the same time. I applied for this so I have control to AWB's code. The bot won't start until we are 100% that mass removal is a good thing to do. Before starting I'll modify the AWB's code not to add Persondata and probably we'll do a new release so that no other editors will add it. -- Magioladitis (talk) 05:41, 2 June 2015 (UTC)
 * We already have consensus for the removal of Persondata. If the addition of Persondata by automated tools hasn't been a breach of COSMETICBOT, then neither should be its removal. Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 07:40, 2 June 2015 (UTC)
 * Hi Andy! You make a good point.  Are there any bots that have been adding Persondata as their primary approved task? GoingBatty (talk) 12:27, 2 June 2015 (UTC)
 * Yes, Rjwilmsibot used to add but not anymore. I already contacted Rjwilmsi about the RfC. -- Magioladitis (talk) 12:57, 2 June 2015 (UTC)
 * Found the approval at Bots/Requests for approval/RjwilmsiBot 4. Thanks!  GoingBatty (talk) 13:33, 2 June 2015 (UTC)

The RFC mentioned above has a section (not an actual wiki markup heading) "Rough plan" which says in part

1. Transfer SHORT DESCRIPTION across to Wikidata. ✅

...

4. Transfer any new data to Wikidata, then remove methodically.

I don't see any agreement to modify the rough plan, so I suppose that is the plan. Will this bot transfer new data to Wikidata, or just "remove methodically." If this bot just removes, how will the part about transferring new data be done? Also, does # 1 mean that if any new data is found, only the SHORT DESCRIPTION will be transferred and other, more suspect, data such as birth and death dates will not be transferred? Jc3s5h (talk) 16:03, 2 June 2015 (UTC)
 * Whoah! I second that concern.  The five-point plant presented at the RfC was expressly conditioned on the "transfer any new data to Wikidata" before the systematic removal is implemented.  This immediate removal without transfer of new input persondata to Wikidata violates the conditions upon which the RfC was approved.  Please adhere to the RfC "rough plan" as presented.  Dirtlawyer1 (talk) 18:23, 2 June 2015 (UTC)
 * I have already suggested that you read the lengthy and detailed discussion of data import under the RfC; and on the pages linked from there, on Wikidata. The RfC was concluded as "deprecate and remove", with no conditions atatched, in the light of that discussion. Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 19:55, 2 June 2015 (UTC)
 * To add, it became apparent during the course of the RfC that no more data would be transferred to Wikidata, all other PD fields having been deemed unreliable. I can't imagine what "remove methodically" might entail. Alakzi (talk) 20:38, 2 June 2015 (UTC) Oh, I see what you mean now, Dirtlawyer1. I agree that the (or a) bot should migrate any descriptions added after PLbot's last run; that would be eminently sensible. Alakzi (talk) 20:56, 2 June 2015 (UTC)
 * Indeed. I didn't just fall off the wiki-turnip truck yesterday.  In addition to the recently added persondata descriptions, I have also raised a concern about the married name variations of female bio subjects listed under alternative names.  Dirtlawyer1 (talk) 21:13, 2 June 2015 (UTC)
 * The outcome of the RfC is "deprecate and remove", not "deprecate and remove with caveats". Andy Mabbett ( Pigsonthewing ); Talk to Andy; Andy's edits 19:55, 2 June 2015 (UTC)

That RFC closed 26 May 2015 (UTC). Wouldn't "deprecate and remove" imply a reasonable period of classic deprecation, possibly with warnings or something, while retaining functionality for some period of time before removal? -- slakr \ talk / 05:27, 4 June 2015 (UTC)
 * See proposal at Bot requests --Francis Schonken (talk) 05:34, 4 June 2015 (UTC)


 * I'm not keen on the sound of "I'll be doing general fixes at the same time." Whatever this bot does, please don't do the other things AWB does at the same time, particularly changing reference positions., if you're doing an AWB update, please remove that instruction. Also, only 35 people supported this RfC. Is that enough to be making mass changes? Sarah (SV) (talk) 20:43, 5 June 2015 (UTC)
 * I won't do any ref reordering. I will wait until there is a stable consensus to remove. There is an active discussion in WP:BOTREQ at the moment. -- Magioladitis (talk) 06:28, 6 June 2015 (UTC)

First batch proposal
I had a look at Bots/Requests for approval/RjwilmsiBot 4: this bot introduced persondata derived from the infobox. Afaics I suppose this automatic removal task would be uncontroversial (apart from maybe cosmeticbot concerns): I see several advantages to carrying out such task now, most importantly some feedback before carrying out possibly more intrusive tasks. --Francis Schonken (talk) 09:19, 6 June 2015 (UTC)
 * yobot removes persondata templates that comply to all of the following:
 * the persondata was created by RjwilmsiBot 4, before the 2014 merge of persondata to wikidata;
 * the persondata template/data has not been modified since its introduction
 * the article still carries an infobox
 * the persondata does not contain any pre-1924 dates
 * Added fourth condition per Bot requests. I still think it useful to operate a first batch of uncontroversial persondata removals. --Francis Schonken (talk) 05:33, 7 June 2015 (UTC)


 * Support - I support the creation of this bot.--BabbaQ (talk) 23:05, 20 June 2015 (UTC)

I think the bot should be approved with no restrictions. -- Magioladitis (talk) 08:07, 5 July 2015 (UTC)


 * So, a couple of still-outstanding issues I'd like to get clarification on:
 * Ensuring no data loss &mdash; if I'm interpreting Bot_requests correctly, I'm seeing some concerns about WP:CONTEXTBOT. For example, ensuring that data contained in the still-present transclusions of  might need to be manually migrated to, e.g., Wikidata or the appropriate infoboxen.  Is everyone certain that things are free and clear, including with the prior actions of prior bots (e.g., RjwilmsiBot 4)?  I'm seeing uncertainty on this.
 * Removing the template &mdash; The concern that the outcome of the RFC is to deprecate and remove the template is clear, but as several have mentioned, a sweeping, automated removal (i.e., the net effect of a template deletion) isn't the same thing, and it could create chaos if it results in data loss. I mean "loss" in the sense of makes it extremely difficult to undo this action once done, as backlinks/transclusions will, in effect, be removed, and the edits will be intermixed with the rest of Yobot's edits, further complicating reverting or parsing disappeared data.
 * RFC substituting for TFD &mdash; TFD is technically the appropriate venue&mdash;not a specialized RFC&mdash;for saying whether a template is safe or desired to be fully unlinked and deleted, hence the difference between "deprecate (and eventually remove)" versus "remove now." I say this not to be bureaucratic, but also because global community consensus is to discuss actual template deletion at TFD, because that's where people will be more likely watching should something be an issue.  The more eyes the better before actual removal, once again, because it'd be a giant pain in the ass to undo.  This obviously ties back in to #1.  Anecdotally, I think someone at one point TFDed one of the  templates long ago, and it was only because of that venue that I knew about it (and obviously the effects it might have had on my bot should it have succeeded).  Similarly, if there are bots or scripts out there that are relying on this still, for some reason, are they going to feel any effects?  In any case, TFD would be the best before automated, mass-removal, most likely.
 * What other general fixes are being done? &mdash; as editors have asked.
 * Granted I could be missing stuff and being over-paranoid, but that's sort of the point before making 1.5 million edits. :P
 * -- slakr \ talk / 19:37, 2 August 2015 (UTC)

on 4 see WP:AWB/GF. -- Magioladitis (talk) 17:30, 3 August 2015 (UTC)
 * As a botop, I feel the level of concern is too great. No disrespect to the botop, and the others involved.  While the idea may be good, it's my impression that as long as there isn't a high level of confidence that this is a good idea, and considering the nature of this task, and how much damage it could possibly cause as a result of a premature removal, I think we should hold off.  We have to hold a discussion somewhere that asks the question, "Should the persondata be removed NOW?"  As botops we are responsible for what kind of damage we can cause by running our high speed programs.  BAG needs to consider the same thing when approving them.  Right now, I'm of the opinion that the risk is higher than the benefit.  No comment on whether removal is warranted or not, but it's definitely not the time to do it now.  I feel BAG has been put in a difficult position here, so these are my 2 cents.—cyberpower  Chat:Offline 04:01, 27 August 2015 (UTC)

&mdash; do we have any sort of update as to what's going on with this or any of the other concerns? -- slakr \ talk / 00:49, 28 August 2015 (UTC)

About the data loss: According the discussions all info about birth/death date/place is already there and there is no evidence that in cases where the values differ that Persondata is more convenient than the other places we store similar info (Wikidata fields, infobox fields). How do you suggest that we solve this problem? We could make create tracking categories before any bot action. Similar to Category:Official website different in Wikidata and Wikipedia. -- Magioladitis (talk) 05:06, 28 August 2015 (UTC)

So we can tweak request to match the one approved for BattyBot and have two bots running. -- Magioladitis (talk) 19:45, 14 September 2015 (UTC)
 * ? Why would we have two bots running? It's not as if persondata need to be removed twice...
 * Recapitulating: above I proposed to start with an uncontroversial batch of removals, to which Magioladitis replied "the bot should be approved with no restrictions".
 * At WP:VPPROP GoingBatty is successful in negotiating the community through one uncontroversial batch after another.
 * Magioladitis declined such negotiative process, and now seems to be willing to ride on the success of a botop who is careful in taking this step by step, and negotiates every step.
 * In sum (1) we don't need two bots doing the same thing, better go help the ongoing process of finding consensus for every step; (2) the step-by-step process works, and GoingBatty garnered more community confidence for it than Magioladitis. (3) Magioladitis seems to underestimate that "tweaking" the request is as well going back to a process he declined, and supposes step by step negotiations, which is something completely different from adding a few words and then let it go loose.
 * For the Yobot 24 BRFA, with all due respect, I think it is time to close it as having gone stale after the recent developments. --Francis Schonken (talk) 03:42, 15 September 2015 (UTC)
 * My negotiation resulted in less than 1,000 removals out of 1.2 million. Everyone's suggestions at negotiating small batches of removal will be appreciated.  If Magioladitis is more successful than I am, I have no problems with him submitting a bot request to remove certain batches.  GoingBatty (talk) 02:04, 16 September 2015 (UTC)


 * I must concur with the above, emphasising that there is no absolute necessity to remove persondata immediately. → Σ σ  ς . (Sigma) 04:04, 15 September 2015 (UTC)

&mdash; Given the discussion, not to mention my own concerns, I can't confidently say that this bot has consensus to run. To help sum up: while there was an RFC that favored deprecation and eventual removal of the template, the latter's timeframe was not clearly established and otherwise appears to present no urgency. Furthermore, several issues are unresolved (or at the very least, not sufficiently reassured to be resolved) with respect to the practical upshot on other forms of automation and/or the process of migrating the now-deprecated template's data to the more appropriate and/or newer representations of the data (whatever that might entail). As such, bulk/blanket removal may serve to unexpectedly confound that process. Because of these issues and e.g., Bots/Requests for approval/BattyBot 47, it seems the likely-less-problematic/likely-less-controversial approach for removal of the template is via piecemeal tasks rather than all at once (for now, anyway). -- slakr \ talk / 21:36, 16 September 2015 (UTC)
 * 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.