[04:05:52] [telegram] I 𝐡𝐚𝐯𝐞 𝐛𝐞𝐞𝐧 𝐢𝐧𝐯𝐞𝐬𝐭𝐢𝐧𝐠 𝐰𝐢𝐭𝐡 𝐭𝐡𝐢𝐬 𝐩𝐥𝐚𝐭𝐟𝐨𝐫𝐦 𝐨𝐯𝐞𝐫 3 𝐦𝐨𝐧𝐭𝐡𝐬 𝐧𝐨𝐰, 𝐚𝐧𝐝 𝐢 𝐡𝐚𝐯𝐞 𝐛𝐞𝐞𝐧 𝐦𝐚𝐤𝐢𝐧𝐠 𝐡𝐮𝐠𝐞 𝐬𝐨𝐦𝐞 𝐨𝐟 𝐦𝐨𝐧𝐞𝐲 𝐰𝐢𝐭𝐡 𝐭𝐡𝐢𝐬 𝐛𝐢𝐭𝐜𝐨𝐢𝐧 𝐢𝐧𝐯� [04:05:52] [telegram] 𝐈 𝐚𝐦 𝐬𝐨 𝐡𝐚𝐩𝐩𝐲 𝐛𝐞𝐢𝐧𝐠 𝐚𝐦𝐨𝐧𝐠 𝐭𝐡𝐨𝐬𝐞 𝐰𝐡𝐨 𝐚𝐫𝐞 𝐰𝐢𝐧𝐢𝐧𝐠, 𝐠𝐚𝐢𝐧𝐢𝐧𝐠 𝐟𝐢𝐧𝐚𝐧𝐜𝐢𝐚𝐥 𝐟𝐫𝐞𝐞𝐝𝐨𝐦, 𝐚𝐧𝐝 𝐚𝐥𝐬𝐨 𝐛𝐞𝐢𝐧𝐠 𝐝𝐞bt 𝐟𝐫𝐞𝐞.. [04:05:54] [telegram] 𝐈𝐧𝐛𝐨𝐱 𝐭𝐡𝐞 𝐚𝐝𝐦𝐢𝐧 𝐟𝐨𝐫 𝐦𝐨𝐫𝐞 𝐝𝐞𝐭𝐚𝐢𝐥𝐬 𝐨𝐧 𝐡𝐨𝐰 𝐭𝐨 𝐢𝐧𝐯𝐞𝐬𝐭. @hendrenbrian [04:05:55] [telegram] https://t.me/+-eHmvXG3vQFhYTg0 [07:41:05] [telegram] Forwarded from unknown: People who give up easily are the people who lose in life, my first experience happens to be a failure but i never gave up because i knew it was going to work out for me to keep trying. I'm smiling today by getting involved with @jamesfxpro [07:41:06] [telegram] https://t.me/+HCRw7RHiTg84NWNk [08:12:34] [telegram] There is also a ticket to be able to provide canonical namespace names for custom ns that was never done I believe [08:13:08] [telegram] https://phabricator.wikimedia.org/T52655 [10:41:34] [telegram] Any idea how can I avoid using user-config.py with pywikibot? I want to pass user name/password when calling script. [10:41:34] [telegram] I found [10:41:36] [telegram] lm = pywikibot.data.api.LoginManager(password=password, site=pywikibot.Site('en', 'osm'), user=user) [10:41:37] [telegram] lm.login() [10:41:39] [telegram] but I still need to create user-config.py because otherwise connection = pywikibot.Site('en', 'osm') will fail... [10:45:30] [telegram] How can I get connection to site without having `user-config.py` file? [10:45:35] [telegram] How can I get connection to site without having user-config.py file? [10:45:53] [telegram] why can't you have a user-config file? [10:47:21] [telegram] Because I would need to either [10:47:22] [telegram] - hardcode login data [10:47:24] [telegram] - create it at runtime which is weird, duplicating config data in multiple places and increasing space where things can fail (re @jhsoby: why can't you have a user-config file?) [10:48:27] [telegram] I am trying to create library - in such case having weird special config files with state lurking in random places is really painful [10:48:49] [telegram] so you want to let arbitrary users run this with their own (bot?) accounts? [10:48:58] [telegram] yes [10:50:26] [telegram] and allow people to do [10:50:27] [telegram] import my_magic_library [10:50:28] [telegram] my_magic_library.handle_images_with_missing_license_templates(user="User", password="xxxx") [10:50:30] [telegram] kind of thing [10:51:03] [telegram] and allow people to do [10:51:04] [telegram] import my_magic_library [10:51:06] [telegram] my_magic_library.handle_images_with_missing_license_templates(user="User", password="xxxx", wiki=somepointer_api_to_be_established) [10:51:07] [telegram] kind of thing [10:51:09] [telegram] in that case i would look into using OAuth instead, and skip pywikibot entirely. [[wikitech:Help:Toolforge/My first Flask OAuth tool]] is a good place to start for that [10:54:16] [telegram] flask requires server to run, right? That is even less viable to me. [10:54:18] [telegram] Though maybe JS-based solution running in browser, client-side is the proper solution... (re @jhsoby: in that case i would look into using OAuth instead, and skip pywikibot entirely. [[wikitech:Help:Toolforge/My first Flask OAuth tool]] is a good place to start for that) [10:54:54] [telegram] This way more people would be able to participate (I am editing on wiki that has 50 000 images with unknown licensing status, !!fun!!) [10:55:37] [telegram] yaay 😆 [10:55:37] [telegram] yeah, i think a user script or preferrably a gadget would be the most user-accessible option in that case [10:55:53] [telegram] Do you need help writing a mass deletion script? :P (re @matkoniecz: This way more people would be able to participate (I am editing on wiki that has 50 000 images with unknown licensing status, !!fun!!)) [10:57:02] [telegram] Well but you definitely should check existing things like AjaxQuickDelete [10:58:14] [telegram] For now? No. [10:58:15] [telegram] But I need a tool that would allow to [10:58:16] [telegram] - scan for images matching some criteria (exclude ones that have licensing templates on file page) [10:58:18] [telegram] - get uploader info for remaining ones [10:58:19] [telegram] - post predefined message on uploader talk page (that references file) [10:58:21] [telegram] - add template to file (depends on date) (re @Thecladis: Do you need help writing a mass deletion script? :P) [10:59:31] [telegram] I tried (as usual) but tools that I have seen either assume direct access to server (rather than using editing API) or are failed to get them to run. [10:59:31] [telegram] And actual coding is quite trivial here, entire complexity is gluing things together. (re @Thecladis: Well but you definitely should check existing things like AjaxQuickDelete) [10:59:42] [telegram] Ukwiki has such a bot more or less I think, but I am not sure if it is active at the moment (re @matkoniecz: For now? No. [10:59:42] [telegram] But I need a tool that would allow to [10:59:43] [telegram] - scan for images matching some criteria (exclude ones that have licensing templates on file page) [10:59:45] [telegram] - get uploader info for remaining ones [10:59:46] [telegram] - post predefined message on uploader talk page (that references file) [10:59:48] [telegram] - add template to file (depends on date)) [11:01:08] [telegram] If anyone is aware of existing tool that can, in browser, client side [11:01:09] [telegram] - list files [11:01:10] [telegram] - get their file page and upload history [11:01:12] [telegram] - make edits to wiki [11:01:13] [telegram] - can be configured to point to https://wiki.openstreetmap.org/ [11:01:15] [telegram] I would deeply thankful [11:01:39] [telegram] https://uk.wikipedia.org/w/index.php?title=%D0%A4%D0%B0%D0%B9%D0%BB:Great_Day_in_Harlem.jpg&diff=prev&oldid=30869495 [11:01:39] [telegram] https://uk.wikipedia.org/wiki/%D0%9E%D0%B1%D0%B3%D0%BE%D0%B2%D0%BE%D1%80%D0%B5%D0%BD%D0%BD%D1%8F_%D0%BA%D0%BE%D1%80%D0%B8%D1%81%D1%82%D1%83%D0%B2%D0%B0%D1%87%D0%B0:%D0%9B%D1%8F%D1%89%D0%B5%D0%BD%D0%BA%D0%BE_%D0%91%D0%BE%D0%B3%D0%B4%D0%B0%D0%BD?diff=cur&oldid=21630999 [11:01:40] [telegram] Such one is actually quite straigtforward to make I think [11:02:01] [telegram] Oh, osm wiki... (re @matkoniecz: If anyone is aware of existing tool that can, in browser, client side [11:02:01] [telegram] - list files [11:02:03] [telegram] - get their file page and upload history [11:02:04] [telegram] - make edits to wiki [11:02:06] [telegram] - can be configured to point to https://wiki.openstreetmap.org/ [11:02:07] [telegram] I would deeply thankful) [11:02:09] [telegram] (right now my plan is to have clunky pywikibot script as I sort of got it to work with this idiotic user-config.py but I worry that noone else will be able to run it) [11:07:24] [telegram] Well if you deal with all 50k in one run it might not be a bad idea. Although on Wikipedia tagging 50k files at once is a sure way to make some new enemies (how do I know this...), perhaps OSM community is more copyright aware though [11:11:25] [telegram] If you end up writing a custom script you would need [11:11:25] [telegram] * https://wiki.openstreetmap.org/wiki/Special:ApiSandbox to see what API can do (mind that the page actually executes the request) [11:11:27] [telegram] * https://doc.wikimedia.org/mediawiki-core/master/js/#!/api/mw.Api for documentation on built in API client [11:11:28] [telegram] * https://www.mediawiki.org/wiki/OOUI to make UI pretty (although that thing is clumpy af, takes 4438437584375 lines of code to do the simplest thing) (re @matkoniecz: If anyone is aware of existing tool that can, in browser, client side [11:11:30] [telegram] - list files [11:11:31] [telegram] - get their file page and upload history [11:11:33] [telegram] - make edits to wiki [11:11:34] [telegram] - can be configured to point to https://wiki.openstreetmap.org/ [11:11:36] [telegram] I would deeply thankful) [11:12:34] [telegram] The script can be made into a gadget that people can enable in their preferences [11:17:36] [telegram] hmm, I was sure I had an account on osm wiki 🤔 [11:19:35] [telegram] Well, there is a plan to have exemption for good faith contributions with irreplaceable images (of events etc). [11:19:36] [telegram] So far deletions are happening and noone complained about them https://wiki.openstreetmap.org/w/index.php?title=Special:Log&offset=&limit=500&type=delete&user=&page=&tagfilter= (but about 1% was done so far) (re @Thecladis: Well if you deal with all 50k in one run it might not be a bad idea. Although on Wikipedia tagging 50k files at once is a sure way to make some new enemies (how do I know this...), perhap [11:20:06] [telegram] note that OSM mapping account and OSM Wiki account are distinct (for now) (re @Thecladis: hmm, I was sure I had an account on osm wiki 🤔) [11:20:46] [telegram] Thank you for this work :) (re @matkoniecz: Well, there is a plan to have exemption for good faith contributions with irreplaceable images (of events etc). [11:20:46] [telegram] So far deletions are happening and noone complained about them https://wiki.openstreetmap.org/w/index.php?title=Special:Log&offset=&limit=500&type=delete&user=&page=&tagfilter= (but about 1% was done so far)) [11:21:31] [telegram] Thing that is the most confusing for me is [11:21:31] [telegram] - how to display fetched images in browser [11:21:33] [telegram] - how to trigger actions when user clicks "uploader should be asked to fix licensing status" (re @Thecladis: If you end up writing a custom script you would need [11:21:34] [telegram] * https://wiki.openstreetmap.org/wiki/Special:ApiSandbox to see what API can do (mind that the page actually executes the request) [11:21:36] [telegram] * https://doc.wikimedia.org/mediawiki-core/master/js/#!/api/mw.Api for documentation on built in API client [11:21:37] [telegram] * https://www.mediawiki.org/wiki/OOUI to make UI pretty (although that thing is clumpy af, takes 4438437584375 lines of code to do the simplest thing)) [11:21:40] [telegram] yeah, I remember attending an OSM related session on 2015 Wikimania just to realise that I am not registered on OSM :D But I was sure that I edited something on wiki too (re @matkoniecz: note that OSM mapping account and OSM Wiki account are distinct (for now)) [18:52:06] [telegram] FYI, this may be of interest - after I nudged a conversation around why commons doesn't allow CSV or database file uploads, SJ Klein made a whole new umbrella bug report to cover a whole lot of other data file formats. https://phabricator.wikimedia.org/T297514 [19:00:00] "why doesn't Commons allow CSV or database file uploads" -> https://commons.wikimedia.org/wiki/Commons:Project_scope#Must_be_a_media_file [19:00:39] the Data namespace is a bit of an exception, but not one the community is particularly looking to widen [19:01:32] [telegram] is 2.1 RDF supposed to be RTF? (don’t want to clutter the Phabricator discussion with this nitpick ^^) [19:57:43] [telegram] Well in such case we need another shared repo for non-media files :) (re @wmtelegram_bot: [irc] "why doesn't Commons allow CSV or database file uploads" -> https://commons.wikimedia.org/wiki/Commons:Project_scope#Must_be_a_media_file) [19:59:32] [telegram] Yay, new project! (re @Thecladis: Well in such case we need another shared repo for non-media files :))