[07:58:14] <Reception123>	 Hi, quick question, where it the "TESTSITE" watermark on test.wikipedia.org defined?
[07:58:31] <Reception123>	 I looked through CommonSettings but couldn't seem to find it
[08:12:29] <Jogo-obb>	 Reception123: https://test.wikipedia.org/wiki/MediaWiki:Vector.css der große Block mit den #p-logo Selektoren
[08:12:45] <Reception123>	 danke Jogo-obb
[08:16:28] <Jogo-obb>	 Reception123: Your welcome ;). Sorry I have not thought about the language before, but I see you have been able to understand it.
[08:17:45] <Reception123>	 It's fine :D I know a little German
[08:19:23] <Jogo-obb>	 Ah ok, I have supposed that, since you answered that fast
[10:15:02] <srdjan_m>	 Is there a way to see a list of local files with the same name on Wikimedia Commons, but where the contents of the local file differs from the one on Commons?
[10:44:31] <MatmaRex>	 srdjan_m: i don't think that exists currently, but it should be easy to write a SQL query for it
[10:53:43] <MatmaRex>	 srdjan_m: https://quarry.wmflabs.org/query/21140 replace 'enwiki_p' with whatever wiki you want to check
[10:54:03] <MatmaRex>	 (it might take a couple minutes to run if you have a lot of local images)
[10:54:52] <srdjan_m>	 MatmaRex: wow, that's pretty neat. thanks!
[11:11:19] <MatmaRex>	 srdjan_m: hmm, quarry seems to be having some issues, but the query definitely works (i get results when running them on analytics databases)
[11:11:38] <MatmaRex>	 i'll run https://quarry.wmflabs.org/query/21141 there
[11:11:58] <srdjan_m>	 hm... ok
[11:13:45] <MatmaRex>	 srdjan_m: https://phabricator.wikimedia.org/P5928 query took 24 seconds
[11:15:03] <srdjan_m>	 MatmaRex: thanks
[11:15:59] <srdjan_m>	 is there an easy way to do that later as quarry.wmflabs.org seems to be stuck?
[11:21:02] <MatmaRex>	 srdjan_m: if you ping me, i can run it again. (i used these databases: https://wikitech.wikimedia.org/wiki/Analytics/Data_access#MariaDB_replicas) for a longer term solution that doesn't depend on me :) , i suggest filing a task about quarry being busted
[11:24:24] <srdjan_m>	 MatmaRex: Hm... could a query be used to find very large images that have a certain license template? For example, this image https://bs.wikipedia.org/wiki/Datoteka:Logo_Afri%C4%8Dkog_kupa_nacija_2006.png is 2000x3228 - it's using {{Sportski logo}} which denotes that it's supposed to be under fair use and much much lower resolution.
[11:27:29] <srdjan_m>	 i found this https://quarry.wmflabs.org/query/18908 in recent queries and don't know if that would work, assuming i changed the categories the templates generate
[12:06:31] <MatmaRex>	 yeah, probably it would
[16:12:26] <Nemo_bis>	 (User rights log) . . Aarontay (talk | contribs) was automatically promoted from (none) to extended confirmed user ‎ (Tag: OAbot [1.0])
[16:12:29] <Nemo_bis>	 https://en.wikipedia.org/w/index.php?title=Special:RecentChanges&tagfilter=OAuth+CID%3A+817
[16:12:51] <Nemo_bis>	 Funny, I didn't know user rights log entries also got tagged :)
[17:41:03] <legoktm>	 Nemo_bis: well you can change userrights via OAuth so it makes sense kind of?
[17:41:51] <Nemo_bis>	 Yeah, it does
[17:42:40] <Nemo_bis>	 But you don't often see it... until a random gorilla wiki creates yet another autopromotion group everybody was longing for
[17:47:40] <quiddity>	 Nemo_bis, I'm not sure if you meant gorilla (as in https://en.wiktionary.org/wiki/800-pound_gorilla ) or guerrilla ( https://en.wiktionary.org/wiki/guerrilla ).  ;-)  (just mentioning for the fun of language)
[17:48:42] <TheresNoTime>	 WMF is really branching out with these gorilla wikis ;P
[17:49:24] <Nemo_bis>	 yeah, the 800 pound gorilla :) but also https://meta.wikimedia.org/wiki/Guerilla_spelling_campaigns of course
[17:50:26] <Nemo_bis>	 (One of the Meta-Wiki deleted pages restorations I'm most proud/happy about.)
[17:56:13] <quiddity>	 ha! that's great.
[18:31:43] <Dispenser>	 What's Aklapper's nick?
[18:32:02] <Nemo_bis>	 andre__
[20:06:23] <Dispenser>	 marktraceur: https://phabricator.wikimedia.org/T156135#3559078 Should I file another bug with the same task description?  Cuz' the example is still a problem.
[20:16:44] <marktraceur>	 Dispenser: The task was to fix detection, so we did that...if you want a maintenance script run, another task would be appropriate I think
[20:17:06] <Dispenser>	 With the same description?
[20:18:10] <marktraceur>	 Dispenser: I'd say just with the request for the script, the detection is fixed so the old title ("...marked as MEDIATYPE_VIDEO") doesn't really apply, because the file wouldn't be marked as such anymore
[20:18:29] <marktraceur>	 I hope that distinction makes sense, I was at about 40% caffeinated when I made it this morning
[20:18:55] <Reedy>	 just needs an sql update running I guess
[20:19:15] <marktraceur>	 I mean yeah
[20:19:22] <marktraceur>	 Reedy: Thanks for volunteering :P
[20:19:33] <Reedy>	 gj reedy
[20:22:36] <Dispenser>	 Why's the subtask marked as high?  Oh well I'll just keep it on the default. >_>
[20:25:17] <Zppix>	 Dispenser: quick question does your disambig tools use pywiki or something i forget
[20:25:49] <Dispenser>	 kinda of, it originally based off it
[20:26:12] <Zppix>	 Dispenser: i see, so you now just write it ourself instead?
[20:26:16] <Zppix>	 Yourself*
[20:28:31] <Dispenser>	 Zppix: pywikipediabot was a piece of shit and as a youngester I didn't know that and figured rewriting the framework so it was fast enough for a CGI script was a good idea
[20:29:40] <Dispenser>	 Mind you Python's come a long way.  Other languages *cough* PHP *cough* still have the shittiest tutorials and advice.
[20:29:44] <Zppix>	 Dispenser: i see, would you advise against using pywikibot to do queries and such and relay them say to file for a webserver?
[20:31:49] <Dispenser>	 I haven't used the latest iteration of pywiki, but the changes make it more suitable than when I started.
[20:32:06] <Dispenser>	 What types of queries are we talking about?  API or Database?
[20:33:20] <Zppix>	 Dispenser: im querying for rev-comments and revision-authors and such
[20:33:25] <Zppix>	 So db
[20:34:44] <Dispenser>	 For database queries, since I have 100 ms lag to my home server, I wrote toolsql.py which basically makes using LOAD DATA easier and fetchall w/ Unicode as default
[20:35:18] <Dispenser>	 I don't think pywiki offers much in database interfaces.
[20:35:41] <Zppix>	 Is the source for that public Dispenser ?
[20:36:51] <Dispenser>	 http://dispenser.info.tm/~dispenser/sources/
[20:37:05] <Zppix>	 Ty
[20:37:36] <Dispenser>	 There some problem I need to work out (e.g. queries with > 4 million results)
[20:37:50] <Zppix>	 Whats the license?
[20:38:09] <Dispenser>	 Grossly Private License
[20:38:15] <Zppix>	 Ah ok
[20:38:35] <Zppix>	 Ill just use it as a guide to rewrite for my own purposes
[20:39:36] <Dispenser>	 There some other issues I need to figure out how to fix, e.g. img_metadata can't be .decode('utf-8')
[20:40:54] <Zppix>	 I think img metadata is encoded in other utf
[20:40:58] <Zppix>	 Iirc
[20:41:22] <Dispenser>	 No its phpserialize and contains binary data
[20:43:21] <Zppix>	 Oh thats right
[21:09:36] <legoktm>	 there's a python library somewhere that implemented unserialize() from php