[06:02:45] [[Tech]]; 112.199.104.126; /* test edit */ new section; https://meta.wikimedia.org/w/index.php?diff=19888135&oldid=19884218&rcid=15043563 [06:02:56] [[Tech]]; DannyS712; test; https://meta.wikimedia.org/w/index.php?diff=19888137&oldid=19888135&rcid=15043565 [08:24:24] enwiki-20191220-pages-articles-multistream.xml.bz2 100%[==>] 16,50G 197MB/s in 2m 5s [08:24:32] * Nemo_bis loves https://ftp.acc.umu.se/mirror/wikimedia.org/dumps/ [08:24:43] I still have the pbzip2 problem though :( [08:40:46] hey Nemo. How are you doing? [09:57:51] apergos: Hi all fine, just have to be more patient with slower decompression ;) [09:58:28] :-) [15:19:59] an anonymous ruwiki user reports that they are affected by block #17774890, yet the blocked range is 2A00:1FA0:0:0:0:0:0:0/33, and the user is using an IPv4 address (which is indicated when attempting to edit) [15:20:42] I remember a similar issue when I just removed the IPv6 range block and placed an equivalent block, and the user could continue editing; how can that possibly be? [15:22:53] edge cases? [15:23:04] Is it a local or global block? [15:23:42] cookie blocks? [15:24:03] Did the anon user ever use an IPv6 address in that range? [15:24:27] that’s a local block: https://ru.wikipedia.org/w/index.php?title=Special:Blocklist&wpTarget=%2317774890&uselang=en [15:25:53] You could ask them to clear their cookies, just in case [15:25:55] Can the IPv6 address possibly be in XFF? [15:26:46] (Sounds the least likely of all options.) [15:42:28] indeed, after clearing cookies, the user can edit again [15:42:33] thanks for helping! [15:43:19] when were cookie blocks introduced on Wikimedia wikis? [15:46:39] np [15:46:40] The option to be able to set cookie blocks on IPs was introduced in 1.32 (according to https://www.mediawiki.org/wiki/Manual:$wgCookieSetOnIpBlock) [15:46:55] But I don't know when they would have been introduced on Wikimedia wikis [15:47:35] (And there might be Wikimedia wikis in which the option is disabled) [16:55:23] hi, I recently had a font added to UniversalLanguageSelector. the font's license is OFL 1.1, like several other fonts there, and as far as I know WMF accepts that as an open license for fonts, but now someone (apparently active on some wikipedia but not part of WMF or involved with UniversalLanguageSelector) is saying the font license has to be CC-BY-SA because OFL is not open [16:55:43] wondering where I can send them to clarify any of this? [16:56:02] I guess https://www.mediawiki.org/wiki/Extension:UniversalLanguageSelector#Adding_fonts [16:59:47] ningu: in general if a font is already in a Debian package then it's much easier to get us (server admins) to actually install it [17:00:05] and you don't have to expect license discussions if it is [17:00:16] well for UniversalLanguageSelector you just commit it directly to the gerrit repo [17:00:26] I don't think there is another way [17:00:27] otherwise i think your point "other existing fonts use the same license already" is a good one. [17:00:53] yeah, OFL is really quite standard, it's like the GPL of fonts (well actually not sure if it's viral or not) [17:01:06] ultimately if you really want to discuss the licensing question i think best you can do is start a mailing list thread [17:02:32] ningu: so Wikipedia says "OFL is a free and open source license.[6][8] The license is considered free by the Free Software Foundation," [17:02:42] "considered free by FSF" is pretty strong [17:02:48] yeah [17:02:57] i would just quote that [17:03:00] so far this is just a person contacting a person who contacted me [17:03:00] and keep going [17:03:08] I asked them for further info/justification [17:03:17] but they are halfway around the world so it may be slow :) [17:03:19] yea, it seems like the burden should be on them if they want to change it [17:03:51] and yes OFL is viral, so it really is like GPL, it's just tailored to fonts and their particular use cases [17:03:59] e.g. embedding in pdfs and such [17:06:17] ningu: We have what we call "wiki layers"... [17:06:30] https://foundation.wikimedia.org/wiki/Resolution:Licensing_policy [17:06:37] maybe this is relevant [17:06:57] There's also 28 other fonts with OFL already [17:07:31] Reedy: yeah that's why I am kind of confused, but I want to be sure to send the right info back over [17:07:41] yea, so the one who wants to question the status quo would have to come up with some more reasoning [17:07:57] I think this is just a person being overly cautious but lacking complete info [17:08:54] I don't even really know who they are [17:10:02] is there a way to get a list of admins on a particular wikipedia site? [17:10:54] ah, Special:ListUsers [17:11:29] Special:CentralAuth/ can be useful [17:11:36] You can find out what they hold where [17:11:42] ningu: https://en.wikipedia.org/wiki/Special:ActiveUsers?username=&groups%5B%5D=sysop&wpFormIdentifier=specialactiveusers [17:12:01] mutante: yeah problem is this person, as far as I can gather, is admin on id.wikipedia.org [17:12:05] but I don't know their username [17:12:23] That doesn't mean they're any sort of authority [17:12:40] How/where did they contact you? [17:12:47] https://id.wikipedia.org/wiki/Istimewa:Daftar_pengguna?username=&group=sysop&wpsubmit=&wpFormIdentifier=mw-listusers-form&limit=50 [17:13:22] they told someone who knows me to contact me, and all I have is a screenshot of their comments on a facebook convo, and I just checked their facebook page and it says they are admin at id.wikipedia [17:13:42] it's a small community of wikipedia folks in indonesia, including the ones in bali [17:13:45] ningu: tell them to come back with a thread on actual wiki :) [17:13:52] haha ok [17:14:42] I just know that, at least culturally, it's best to get out in front of these things -- people want to feel included and if I do nothing it could create resentment. so I'd rather not just ignore it entirely [17:15:02] You can reach out to the WMF legal team too [17:15:13] They're good with licensing stuff, and have people who definitely know what they're talking about [17:15:23] I mean, that seems a bit much given that this has already been decided by people who know more than me [17:15:28] and it's clearly stated [17:15:53] I think asking this person to post in an on-wiki thread makes perfect sense [17:15:57] I mean if you can't get them to listen/understand etc [17:16:01] and we can just go from there [17:16:29] i think the first step has to be that person explaining themselves (what do they want to change, why) in some public place [17:16:42] if after that there are still open questions.. then they can ask WMF legal if they want [17:16:42] like I said I think it's just someone who heard about what's going on and wants to share their two cents, and they noticed the font license used to be non-open, so wanted to warn us -- although they didn't see that the license has since been changed after I contacted the author about it [17:17:14] they found one of the transcripts of my conversations on here where I said the old license may be an issue but I'll try contacting the author to change it [17:18:10] it's not clear to me (yet) if they'll really cause problems or just want to chime in and make sure it's all ok [17:19:05] ok I just realized I do know who it is [17:19:14] I mean, their on-wiki name [17:19:39] ningu: i think it's very likely it's "just wants to chime in" [17:19:54] yeah, they wrote in support of my project grant [17:20:25] so yeah, I don't want to escalate, I just want to provide clear info [17:21:59] I'm actually very happy that the font author and people in bali have been ok with the open licenses for their work [17:22:21] should say free I guess, not open [17:23:10] ningu: another option could be to discuss on Talk:SIL_Open_Font_License [17:23:19] (en.wp) [17:23:34] already has some "GPL compatible?" section [17:39:44] <_joe_> Nemo_bis: not sure I can answer your question on twitter [20:28:17] _joe_: Ah sorry didn't mean to ask any secret. :) Once upon a time one could just tell from the public charts in AMSIX anyway [20:28:49] <_joe_> yeah sorry, I did the query but on a tool under NDA [20:29:05] <_joe_> so I'm not sure how free I am to divulge the results [20:40:19] https://grafana.wikimedia.org/d/000000607/cluster-overview?orgId=1&var-datasource=esams%20prometheus%2Fops&var-cluster=cache_text&var-instance=All&from=1582836007361&to=1584045607362&fullscreen&panelId=84 [20:41:14] there's a spike in traffic through esams that correlates pretty well [20:44:41] <_joe_> AntiComposite: oh wow [20:44:44] <_joe_> you're right [20:50:08] <_joe_> AntiComposite: it seems to be something we did on our side [20:50:17] <_joe_> see eqsin https://grafana.wikimedia.org/d/000000607/cluster-overview?orgId=1&from=1583806417183&to=1583863938922&fullscreen&panelId=84&var-datasource=eqsin%20prometheus%2Fops&var-cluster=cache_text&var-instance=All [20:58:16] yeah, looking at the dashboards for the caches themselves I don't see any significant anything [20:59:25] hmm...changes to restbase around 2020-03-09 12:00:00? [20:59:51] <_joe_> I don't think so [21:06:26] ¯\_(ツ)_/¯ definitely looks like someone did something to cause more traffic with the same number of requests