Commons:Bots/Requests/AnankeBot

AnankeBot (talk · contribs)

Operator: Nemo bis (talk · contributions · Statistics · Recent activity · block log · User rights log · uploads · Global account information) (assign permissions)

Bot's tasks for which permission is being sought: Import some Flickr albums/photostreams which would disappear at the beginning of February 2019 and which have adopted a free license, as part of the Flickr preservation campaign where I'm writing to thousands of photographers.

Automatic or manually assisted: Mostly automatic

Edit type: Occasional runs

Maximum edit rate (e.g. edits per minute): No more than 120 uploads/minute in the worst case, I expect

Bot flag requested: (Y/N): Yes

Programming language(s): Python: a slightly adapted flickrripper.py (to add categories)

Nemo 15:16, 1 January 2019 (UTC)

Discussion

• Are there any filtering planned? For example, Commons:Project scope, Commons:Derivative work, Commons:Freedom of panorama, etc. --EugeneZelenko (talk) 15:32, 1 January 2019 (UTC)
  • For now I'm doing albums manually, choosing the streams which contain almost only suitable photos, although the occasional less-than-perfect photo may slip in (e.g. a selfie in an album of hundreds photos of a place). These users have already been selected as suitable/interesting by the community across the years by importing and keeping their photos in the past. Nemo 15:41, 1 January 2019 (UTC)

    I think will be good idea to create accompanied project for this task and at least needs review category. --EugeneZelenko (talk) 15:26, 2 January 2019 (UTC)

    • I'm adding the {{Check categories}} where no specific category is found. I'm also going through the generic categories after the fact for any bird's eye corrections. This page and Commons talk:Flickr files can serve as project page. Nemo 15:38, 2 January 2019 (UTC)

      I didn't mean categorization. Just service category which indicate that file was not reviewed by human. --EugeneZelenko (talk) 14:56, 3 January 2019 (UTC)

      I can do that, but what such categories are available to use? If you mean for copyright/scope issues, I think I'll manage to go through the categories after the upload and mark most problematic cases for deletion myself. Nemo 17:55, 3 January 2019 (UTC)

      I meant maintenance category dedicated for this task. If bot will upload a lot of files it'll be better to ask other people help, so in this case such category is better solution. --EugeneZelenko (talk) 15:16, 4 January 2019 (UTC)

      I don't currently expect the bot to upload such a significant amount of files as to require an entirely new maintenance system. It will be comparable to or lesser than previous imports from Flickr.

      To be clear: I'm happy to create a new category if at least one actual person appears who actually wants to use them, but I'm weary of inventing new categories for purely hypothetical users. I prefer to use existing systems which integrate in existing workflows and "check categories" does, because it works very naturarly with people who sort files by using HotCat, cat-a-lot and related maintenance gadgets/scripts. I've yet to find people who go through Flickr files by using other ad hoc categories. Nemo 13:07, 5 January 2019 (UTC)

      Categorization and copyrights issue or project scope a little bit different problems. --EugeneZelenko (talk) 14:49, 6 January 2019 (UTC)

      Indeed, which is why I'm weary of inventing a new maintenance apparatus with no users which would replicate an existing system (tracking categories) already used in a different way. If someone builds it and uses it, I can adopt it. Until then, I'll stick to what exists, i.e. categories for categorisation work (through which most of the time scope and copyright issues are typically found) and manual high-level checks after the fact. Nemo 18:10, 6 January 2019 (UTC) P.s.: I'll note that slowing down this kind of bot requests only encourages people to perform imports with their main accounts (where it's easier to get upload_by_url and ratelimit excemption), which makes them less transparent.

      • @Nemo bis: Signing your posts on talk pages is required by Commons:Signatures policy. To do so, simply add four tildes (~~~~) at the end of your comments. Your user name or IP address (if you are not logged in) and a timestamp will then automatically be added when you save your comment. Signing your comments helps people to find out who said something and provides them with a link to your user/talk page (for further discussion). Thank you.   — Jeff G. ツ _{please ping or talk to me} 14:55, 13 January 2019 (UTC)

• I don't think this should be done by a bot account, but I do think it should not be done by a bot account before approval and without autopatrol. --Krd 06:52, 12 January 2019 (UTC)
  • I could do the current uploads from my main account or from an account without "bot" in its username, as is customary on Wikimedia Commons, but I prefer to keep my contributions smaller. The absence of flags is easy to fix. Nemo 11:37, 13 January 2019 (UTC)
• I tend to agree with Krd. There is a serious risk of massively importing copyvios, also given some recent issues with WMI projects you promoted. There is a way to check that the imports are not made from "flickrwarshing" accounts or other sorts of copyvios? --Ruthven (msg) 20:52, 30 January 2019 (UTC)
  • I'm checking the accounts manually. There wasn't any issue with the books by w:Emilio Salgari (who died in 1911). We got the earliest editions available at the national central library, knowing that some pages with illustrations by recent authors would have to be removed. Nemo 17:23, 31 January 2019 (UTC)

    @Nemo bis: Who is this "We"?   — Jeff G. ツ _{please ping or talk to me} 17:29, 31 January 2019 (UTC)

    I and the uploader of those files. Nemo 18:10, 31 January 2019 (UTC)

    @Jeff G.: This issue was quickly solved without problems. @Nemo bis: My concern is that such files in copyvio (and other that you might upload with the bot) were not "filtered" at the source. Mistakes can happen, but here we want to be assured that there will not be massive uploads of copyrighted material or other stuff that can damage the project. This is why, some control on each account (and possibly on each file - e.g. for FOP) should be performed by human. Ruthven (msg) 08:59, 1 February 2019 (UTC)

I currently see some consensus this should not be done by bot. If I'm mistaken, please advise in detail. Thank you. --Krd 17:20, 15 February 2019 (UTC)

Closing as declined per above. --Krd 13:38, 7 March 2019 (UTC)