Wikidata:Requests for permissions/Bot/Dexbot 3
The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section. A summary of the conclusions reached follows.
- No consensus to run this task. The Anonymouse (talk) 19:44, 16 December 2013 (UTC)[reply]
Dexbot (talk • contribs • new items • new lexemes • SUL • Block log • User rights log • User rights • xtools)
Operator: Ladsgroup (talk • contribs • logs)
Task/s: Removing statements of deleted properties
Function details: When a property will be removed I'll run bot in order to remove statements in items --Amir (talk) 23:57, 13 June 2013 (UTC)[reply]
{{approved}}
Operator is trusted and competent. Legoktm (talk) 00:41, 14 June 2013 (UTC)[reply]
- Oppose: you'll often need to check properties on a case-by-case basis, and such actions would cause loss of data. @Legoktm: this is not uncontroversial at all, please reopen. --Ricordisamoa 01:54, 14 June 2013 (UTC)[reply]
- Also, I notice Dexbot has already mass-removed "uncle" from items, without ever checking the existance of other kinship properties. --Ricordisamoa 01:57, 14 June 2013 (UTC)[reply]
- The bot was tasked with removing the property. It removed the property. That's exactly what I asked for. The existence or lack of other kinship properties is immaterial, because they're all imported at this point. If they're missing it's because we really haven't done the importing we need to. Sven Manguard Wha? 02:25, 14 June 2013 (UTC)[reply]
- But this can't be extended to all properties beforehand. --Ricordisamoa 02:40, 14 June 2013 (UTC)[reply]
- The bot was tasked with removing the property. It removed the property. That's exactly what I asked for. The existence or lack of other kinship properties is immaterial, because they're all imported at this point. If they're missing it's because we really haven't done the importing we need to. Sven Manguard Wha? 02:25, 14 June 2013 (UTC)[reply]
The task in uncontroversial and simple, when we reach a consensus that a property should be deleted and data should be removed i run the code and remove all of that. I don't remove every property I like (Wikidata is not my property) for uncle, see WD:PFD of P29 my bot massively removed property after closing the PFD as "delete"Amir (talk) 04:02, 14 June 2013 (UTC)[reply]
- It is not true that there all imported at this point. Per 2013-06-10 there was 28 statements using P:P139 (aunt) and none of these had references. There was 575 statements using P:P29 (uncle), and only 9 these had references. The references was:
reference statements Imported from German Wikipedia 3 Imported from English Wikipedia 2 Stated in The Bud Collins History of Tennis (2nd ed.) (Q13415851) 2 Stated in La Préhistoire des Capétiens (Q13422577) 1 Stated in 100 years of Wimbledon (Q13410659) 1
- I was the person that closed the discussion, and I was the person that asked for a bot to come in and delete the statements using the property. Just because you apparently don't agree with the arguments made for deletion (which is something that you needed to have said at the PfD) doesn't mean that there isn't a consensus. Dealing with kinship properties has been tremendously complicated, and suffers from the problem that lots of people have opinions, and those opinions tend to differ, but when the issue was brought to RfC, and even when it was linked in the watchlist notice (it isn't anymore, I pulled it a few hours ago) very few people bothered to try and forge a solution. The PfD itself was left open for two weeks. I'm not entirely sure what you wanted as a result, but in the absence of anyone actually stating their alternatives, and getting support from them, I worked with what I had; five comments in favor of deletion plus the nomination itself. Sven Manguard Wha? 05:36, 14 June 2013 (UTC)[reply]
- I do agree with the arguments made for deletion of the properties, and thus felt no reason to add another "me too" at the PfD. And I do agree that there is consensus for deletion of the properties. That I not agree about is that there is consensus for deletion of the statements using the properties before the data was restated with other properties. Byrial (talk) 06:08, 14 June 2013 (UTC)[reply]
- I was the person that closed the discussion, and I was the person that asked for a bot to come in and delete the statements using the property. Just because you apparently don't agree with the arguments made for deletion (which is something that you needed to have said at the PfD) doesn't mean that there isn't a consensus. Dealing with kinship properties has been tremendously complicated, and suffers from the problem that lots of people have opinions, and those opinions tend to differ, but when the issue was brought to RfC, and even when it was linked in the watchlist notice (it isn't anymore, I pulled it a few hours ago) very few people bothered to try and forge a solution. The PfD itself was left open for two weeks. I'm not entirely sure what you wanted as a result, but in the absence of anyone actually stating their alternatives, and getting support from them, I worked with what I had; five comments in favor of deletion plus the nomination itself. Sven Manguard Wha? 05:36, 14 June 2013 (UTC)[reply]
- Oppose – You should not approve a bot to remove all statements from all deleted properties. There may be cases where data should be restated using other properties and/or checked manually. And you should certainly not approve this after less than one hour before anyone had time to comment. Byrial (talk) 05:15, 14 June 2013 (UTC)[reply]
- Perfectly agreed, I was preparing a huge comment but Legoktm conflicted me... --Ricordisamoa 05:46, 14 June 2013 (UTC)[reply]
- Well... efff. It looks like we screwed this one up. Any ideas on how to resolve it? Obviously this task shouldn't be run again until the bot is updated to handle removals in a better fashion. Sven Manguard Wha? 17:26, 14 June 2013 (UTC)[reply]
- Support If it should be replaced by another property, we need not use this particular bot. The deleted data can always be found in the page history anyways.--Jasper Deng (talk) 05:06, 15 June 2013 (UTC)[reply]
- Info I've just reopened the discussion ( On hold). Vogone talk 16:54, 15 June 2013 (UTC)[reply]
- Comment My sincere apologies for speedily approving this, I didn't realize it wasn't as simple as it seemed. We had discussed this on IRC and the implementation seemed fine so I went ahead and approved it. Thanks for catching and realizing my mistake. Now, for a way forward. Do we need to re-instate the removed data? What properties need to be re-added? Legoktm (talk) 20:14, 20 June 2013 (UTC)[reply]
- It might be an idea to try to find out who inserted the P29 and P139 claims, notify them about the deletions and tell them that they can give the same information with the father/mother + brother/sister properties. There is a list with the claims at June 10 at User:Byrial/Uncles. Addition of claims with the wbsetclaim API call can relatively easily be found in database dumps (or in the revision table for those with access to a replicated database) from the autogenerated edit summary. Byrial (talk) 00:09, 21 June 2013 (UTC)[reply]
@Ladsgroup, Legoktm, Jasper Deng, Byrial, Sven Manguard, Vogone: I propose we close this discussion as unapproved because of the disagreements. --Ricordisamoa 03:25, 7 December 2013 (UTC)[reply]