[12:44:33] Is there a way to check the most populated category? [12:52:03] Check? [13:28:25] Special:MostLinkedCategories [14:00:25] sjoerddebruin I am looking for most used [14:00:55] ah I see now [14:00:59] member means its in it [14:01:00] ty [14:01:07] yeah, it could have a better name imo [14:59:20] hmm I am writing a querry to find the nth file [14:59:44] page_namespace = 6, but I want to ignore non file pages [15:00:07] sjoerddebruin suggestions? [15:29:12] I have https://quarry.wmflabs.org/query/14418 [15:30:39] sjoerddebruin is that right? Is there a more efficient way? [15:31:14] That won't work will it? [15:31:36] Maybe it will.. [15:55:18] Reedy it kinda did [15:55:24] I am unsure if I am only counting the files. [15:55:47] since you can create non-redirect file: pages [15:56:18] how can I check if a binary file is associated with a file: page? [15:56:38] join on the image table? [15:58:49] can blocked users still send public thanks for edits? [16:03:24] probably not, but it's not impossible they can. let's find out [16:04:43] Dragonfly6-7: they definitely can't [16:06:07] ah, good. [16:06:30] hm [16:07:09] MatmaRex, check your msg [16:43:34] Reedy can you help a little, I am a bit lost :( [18:11:45] could someone at least tell me which table handles binary files? [18:15:48] "handles"? [18:18:34] I want to link file: namespace pages with binary files. [18:18:51] so that I get a list of pages that pair with binary files. [18:29:35] ToAruShiroiNeko: information about file uploads is in the image table (and oldimage for older versions, and filearchive for deleted uploads) [18:30:20] i think you can join on page_title=img_name [18:30:42] hmm [18:30:49] that should do the trick [18:31:38] to give you an idea, the 35,000,000th file was uploaded to Commons today, we (well, probably just me) want to find out what the file was. I've coerced ToAruShiroiNeko into writing doing the quarry stuff [18:33:49] MatmaRex https://quarry.wmflabs.org/query/14418 [18:33:54] this query is dismally slow :/ [18:34:04] not suprising since its trying to pair 35,000,000 files [18:34:30] *rows [18:48:44] MatmaRex what would your query be for the 35 mil file, taking deletions into account? [19:12:55] ya querry was killed :( [19:38:51] Reedy do you have a suggestion? :/ [19:57:52] NotASpy: ToAruShiroiNeko: given that files are being deleted all the time, i don't really think it's possible to identify which one was 35,000,000th [19:58:21] i mean, yes, at some point there were 34,999,999 files, and another one was uploaded. but it's going to be difficult now to identify which one that was [19:59:07] * NotASpy nods [19:59:18] andre__: https://gerrit.wikimedia.org/r/#/c/324221/ is about twice as much work as I was expecting for one task instance fwiw [19:59:31] how do we do it for articles on en.wp, for example ? [19:59:40] you could probably analyze the upload and deletion logs, somehow, and you would get some file, but it wouldn't really meaningfully be the 35 millionth one [20:00:35] NotASpy: back when pl.wikipedia was about to reach 1 million of articles, i remember there was a bot watching both the counter and the new pages trying to identify the millionth one. it's a lot harder to do restrospectively though [20:00:45] legoktm: Yeah I was also surprised to see the number of line changes. Was wondering first if the student did a mistake :) [20:00:47] NotASpy: honestly we look at the page_ids around the time the site stats counter went over the milestone, and then randomly pick one that's close enough [20:01:42] legoktm: so if things turn out to be way more complicated than expected feel free to let the student pass, with an optional sentence explaining the obstacles of estimation in software development :) [20:02:06] alright :) [20:02:50] I thought about trying to work out which file tripped the counter over 35,000,000 and simply then ignoring if a hundred deletions then left us with 34,999,900 files. [20:06:03] NotASpy: if you take the current number of files, and then go backwards through deletion and upload logs, that can probably be done… [20:10:36] yeah. It's probably a bit late now to do it, just the 5300+ live files and all the deletions from 14:45ish UTC onwards to review. [21:59:07] bblack (or someone else, but you wrote the incident report), if a user (through OTRS) is reporting (four days ago) that Safari 4.05 (OS 10.5.8) is complaining about Wikipedia's certificate being untrusted, and that when he checks "always trust," images are no longer displayed, could that be related to the GlobalSign incident or another known https-related issue? [22:05:58] is that MacOS 10.5.8? [22:06:08] Must be [22:06:37] That sounds an ancient version of Safari too.. [22:06:44] Yeah, he wrote "OS 10.5.8", but the subject line reads "Mac OS 10.5", so I guess so [22:08:41] Old versions of OS' and Browsers can give all sorts of weird issues [22:08:51] Says this was the second time in the past month he's getting this "certificate untrusted" message on Wikipedia. [22:24:36] pajz: did the problem already go away for the user? [22:25:24] bblack, after the first time he said it went away after about 10 days [22:25:36] there was a window of time for a few hours circa Nov 22-ish (I'd have to look up the window) when a very rare subset of users would've seen the same globalsign issues as the original incident [22:26:17] (very rare being: Safari users who observed the original issue back when it happened, and then were also unlucky I guess) [22:26:38] the original issue was... I think about a month before that? [22:27:33] mid-october, yes [22:30:08] Do you know of any way the user could verify if his problem is in some way related to the Nov 22 issue or would that have disappared immediately after it occured? [22:31:36] it should have dissappeared afterwards, but the window was a few hours long [22:31:45] certainly shouldn't still be having any issue today related to that [22:32:54] for users that did have that issue temporarily, this was a fix for it: https://support.globalsign.com/customer/portal/articles/1353318 [22:33:01] but even that fix shouldn't be necessary at this point [22:33:29] otherwise I guess standard advice about updating browser/os patchlevels and making sure everything's up to date [22:33:48] and try disabling "security" software that tries to proxy TLS connections :P [22:33:59] (which is quite common, and quite commonly broken) [22:34:15] the kinds that claims to have content filtering to prevent malware or block porn or whatever [22:35:40] Ok. I see. Thanks for your help, bblack and Reedy!