[00:00:10] it's fun that it's " [00:00:10] How Wikitia helps you [00:00:25] From what i understand, they only used the code review feature, they never used the task tracking part of it [00:00:26] but it actually has "pay us to edit the content yoourself" [00:00:44] Honestly, weird wiki project with verified only accounts shilling crypto [00:00:50] I'm surprised sanger isn't behind it [00:01:01] he can only do one of those at a time [00:01:10] That's very surprising that he's not behind it [00:01:21] I thought it was a given [00:01:30] bawolff: wanna know how I learned about the existence of that site? through getting a contact request on .. LinkedIn [00:02:03] ... [00:02:13] LinkedIn: The only site on the planet that manages to be creepier than facebook [00:02:20] that WLM logo is hotlinked from us [00:03:11] https://lists.wikimedia.org/hyperkitty/list/wikitech-l@lists.wikimedia.org/thread/4N7QWHBHZBRYSPQGGG5E4BZBYAYKKY4C/ about time [00:03:17] Gosh wouldn't it be terrible if we served a different image to that referrer [00:03:21] That was always such a weird behaviour [00:04:22] the person who says they are a "verified editor" at wikitia and connects WMF emplyees.. offers the following services: Content Marketing, Content Strategy, SEO, Social Media Marketing.. and .. Instagram Verification [00:04:38] rofl [00:06:20] Also "wikiedata.com" [00:06:26] Doesn't seem to work for me [00:06:40] But linked in the left sidebar [00:07:22] For all your verified unrefenced data :P [00:08:01] * TheresNoTime didn't just check to see if that was available.. [00:10:11] it's taken and behind cloudflare [00:11:11] wonder why they got into wikidata? https://www.semrush.com/blog/seo-professionals-how-to-get-started-with-wikidata/ [00:12:38] pfft and yet I'm not allowed a wikidata item [00:12:45] "Create at least one backup account. " [00:14:11] gotta go, cu guys. this channel got an activity spike today.. :) thanks Dreamy_Jazz [00:14:29] :D [00:14:40] yes, its nice to see more activity on the main public irc channels. Sometimes it can feel like they are dying [00:14:43] o/ [00:15:23] Could always restart the weekly code review "office hours" here [00:15:31] (pls.) [00:15:44] TheresNoTime: Just convince yaron to interview you. Instant wikidata item :P [00:15:46] I'd love to have that tbh [00:15:52] Not a bad idea tbh [00:16:13] Having a weekly code review time would be great to get things through that need back and forward between those reviewing and submitting [00:16:44] This might just be the best channel for it.. not too noisy [00:17:25] I think if we did it again though, we should maybe make the requirement that it be for your own patches only. I think last time around there were people who were trying to get patches through that weren't there own, which i understand, but it didn't work too well [00:19:30] lol, one more comment I had to leave here: https://page.wikitia.com/choose-plan/ they are taking money for editing [00:19:41] "choose your plan" after you click on edit.. hahahaa [00:20:02] and now out again :) [00:21:06] rofl, $100 [00:21:15] That's ambitious for a site nobody has heard of [00:22:03] I wonder if the BMW testimonial is real [00:22:10] Is this the truth social of wikis? [00:22:21] > These days, mostly hate crime content is highlighted and not the positive work of individuals. [00:22:27] >.> [00:26:49] bawolff: https://www.businesstoday.in/latest/corporate/story/bmw-india-rocked-by-sudden-demise-of-ceo-rudratej-singh-255975-2020-04-20 [00:26:56] he died in 2019 [00:27:29] That's actually pretty cold [00:27:40] To fake a testimonial of a dead man [00:30:36] I wonder if we could have a wiki page where people can list the patches they would like to be reviewed, and then a jouncebot-ish bot could announce them here weekly at the start of the review hours [00:33:05] End up with a decent list of patches wanting attention the rest of the week too.. [00:33:08] I am a bit skeptical if adding the need for realtime / to do reviews at a specific time of day actually improves the likelihood of people reviewing. we tried this for puppet patches [00:34:28] Perhaps if it results in one extra patch reviewed a week, it's still a net positive? [00:35:00] I kind of like it as an idea of connecting people in order to figure out what the next step is [00:35:17] Not neccessarily actually reviewing the patch [00:35:18] it lead to an incentive to downplay changes as more trivial than they might be to get them into the window.. or people feeling more pressue to merge things they would not normally merge [00:35:51] then you end up with some rule that the window is only for things that are already reviewed and you still have the review problem [00:36:19] but could be different for you since you talk about mw core [00:36:26] Or maybe to have a sync conversation about the patch so that all parties understand what needs to happen next [00:36:38] I mean, it definitely did not really work last time we tried it [00:37:04] Yeah I'm coming into this with my trademark rose tinted glasses [00:37:16] Boundless optimism, etc [00:37:52] But at other jobs i've had, i definitely did find that synchronous code review for tricky patches was helpful. Maybe not for normal ones though [00:38:19] although that was more like over zoom [00:38:31] it makes me think why that same conversation can't happen on the review tool itself.. because under normal circumstances _removing_ realtime should make it more effective [00:38:49] I feel like, and I have zero evidence to back this up, the sorts of patches which would appear in those office hours are the sort that need discussion rather than a +2 stamp [00:39:27] Which makes it hard, because to have an effective discussion you have to be familiar with the patch, and if it needs a discussion it probably means it will take more than 15 min to get familiar [00:40:04] wiki talk pages have discussions without the need to all speak at the same time.. why cant it be that way on gerrit [00:40:14] mutante: Up to a point, but context switching and long delays have a very real cost. Someone review your patch, you address it, and then 2 weeks later they look at it again. By that point both of you have forgotten what it is all about [00:40:37] Wiki doesn't have pre-commit review at all though :) [00:40:38] (otoh I'm sure there's also instances where all a patch does need is a +2) [00:41:20] Dreamy_Jazz: for example, how are you finding getting CU patches reviewed? [00:42:40] bawolff: understood.. yea.. I would argue the real issue is the "2 weeks later they look at it again" part there.. assuming people receive email from Gerrit .. and they would not find it acceptable to wait 2 weeks for an email reply.. [00:42:52] Much better with z.abe often reviewing patches. I've noticed it often is a back and forward with my patch and then reviewing a patch. [00:43:24] Which means the actor and comment migration has been going quickly. [00:43:24] And really, 2 weeks is not even that bad a turn around time for mediawiki patches in underserved areas [00:43:28] back and forth means review is working well [00:44:46] Like, I made Extension:QuickInstantCommons an extension rather than core almost entirely because I got disheartened on turn around time in code review [00:45:13] Plus I can use for some of my recent patches the justification that there are several security tickets that depend on the patch. [00:45:27] e.g. the big one currently in review [00:45:28] But lately I've been doing work on some extensions that are kind of less obscure area, and i found some of my patches reviewed surprisingly fast [00:45:49] the actual issue seems that nobody is getting paid to do it [00:45:53] ^ [00:45:55] Dreamy_Jazz: That was my favourite part of working in security back when i used to. Nobody can say no security ;) [00:46:11] Maybe this is a solution in want of a problem then.. [00:46:46] Could always send out a survey, a la "do you feel like your patches are reviewed in a reasonable amount of time?" [00:47:15] Ultimately yes. But at the same time there's a lot of things people do without being paid. It raises the question of why is code review so special [00:48:02] But tbh, I like working on CheckUser because I see the need for it. [00:48:19] As when I run a check on enwiki I can notice the difference [00:48:36] When are you doing global CU [00:48:46] TheresNoTime: you will be happy to hear we are about to send out such a survey [00:48:48] My personal theory is that code review relies almost entirely on social connections, and social connections in the volunteer community don't overlap with areas of work, like they do when you work for a WMF team and everyone on your team is working on the same code base and dependent on code review to make process [00:48:49] MediaWiki Foundation should hire you [00:48:52] TheresNoTime: funny you should mention that...we're gearing up for another https://www.mediawiki.org/wiki/Developer_Satisfaction_Survey/2021 :) [00:49:16] and that is nearly verbatim one of the questions [00:49:23] re: why code review is special, my pet theory lately is that code review just sucks. [00:49:36] brennen: I can buy that theory [00:49:49] reviewing code really well is often at least as much as effort as writing the change under review. [00:50:05] we already know the result of that survey question though [00:50:13] Glad to hear a survey is on the way :D [00:50:18] you have to model not only the code under change, but whether the change itself is a) a good idea and b) working as intended. [00:51:00] I agree that WMF should hire Dreamy_Jazz. He's doing great work on stuff that is critical to the well running of wikipedia :) [00:51:11] revised model: a) what's the worst that it can break b) if not appservers.. yolo [00:51:12] unless you actually meant the "MediaWiki foundation" [00:51:20] mutante: :) [00:51:28] brb starting the mediawiki foundation [00:51:52] Global CU is a an interesting one - Tbh I'm not sure that there would be the support for it cross-wiki and it will be a lot of work. [00:52:18] Dreamy_Jazz: Who needs global check user when you have logstash? (That's a jk in case anyone quotes me out of context) [00:52:43] Thanks b.awolff :D [00:52:57] Just place a reason field infront of logstash :) [00:52:57] bawolff: I actually meant MediaWiki foundation :p [00:53:03] the real answer is: only post-merge review allowed, just like on-wiki. [00:53:17] the infrastructure anyone can edit [00:53:22] :D [00:53:23] "MediaWiki foundation" is a hand-wavy collection of unclear ideas that lacks leadership talent to make it into a real idea [00:54:03] thcipriani: Bring back Special:CodeReview ! [00:54:13] i dunno, it meets the "naming must be incredibly confusing" criterion for things operating in this space. [00:54:19] Dreamy_Jazz: just so *I* don't get quoted, I'm joking ref Global CU :p [00:54:31] Ah. I see. [00:54:42] Wasn't going to quote, but good to hear [00:54:46] I had no problems getting my code review when it was just chad doing post-merge review on literally the entire code base once every six months [00:54:55] Sucked to be chad though [00:54:58] (iirc a lot of people are against it) [00:55:08] Yeah.... [00:55:20] Looking at the task about loginwiki I could see that [00:55:44] If fandom need it they can make it themselves [00:58:34] The developer satisfaction survey is pretty neat :) I'm glad that exists [00:59:33] In a general point, if anyone needs something done in the CheckUser extension feel free to make a patch and ask me to review. [00:59:59] I'm kind of surprised that the local dev experience is lower than code review [01:00:18] Although maybe im happier with it because i don't use docker or vagrant [01:00:50] being able to unbreak your own thing is probably a big factor :) [01:01:12] It took me a while before I was able to get used to Docker and recently vagrant [01:01:23] The docker compose setup for local development is pretty painless tbh.. [01:01:26] But I do like the containerised nature of docker [01:01:42] For postgres it's currently broken, but I'm going to fix the patch for that. [01:01:50] *fix the documentation [01:01:55] I feel like vagrant was pretty terrible. I hear docker is a bit better [01:02:11] But i always feel like both of those systems kind of make the already easy parts easier, but the hard parts much harder [01:02:13] Vagrant is crap [01:02:16] :) [01:02:20] but i may be in the minority [01:02:26] I'm managing a vagrant wiki as a psuedo-beta wiki [01:02:42] Specifically for CheckUser changes [01:02:56] I would have no idea how to do it without the guide [01:02:58] "Are you satisfied with beta" <-- Can't be unsatisfied with something you never use ;) [01:03:36] It's okay it's getting phased out (ahahahaha ahaha ah ha.) [01:04:00] * thcipriani weeps quietly [01:04:05] :D [01:04:24] Hey at least mw-on-k8s is looking really promising! [01:04:33] What does your patch not working on beta tell you? It tells you beta is currently broken [01:04:35] it is? the latest message I heard was first that we prioritize it and then that we decided to postpone that [01:04:55] mutante: that's the joke 😌 [01:05:25] "we are going to prioritize this" is WMF code for it is being deprecated soon [01:05:25] serving "live" (test2wiki) traffic, that's something [01:05:30] See also flow [01:06:09] https://isbetabroken.com [01:06:29] TIL [01:06:35] that's amazing :) [01:06:52] That is a great site name [01:06:59] to be honest I was braced for a page that said "yes" [01:07:07] Same here [01:07:08] https://butreally.isbetabroken.com ... [01:07:18] HAHA [01:07:24] hahaha [01:07:34] define "work" [01:07:38] production-like [01:07:59] As in, it has MediaWiki running, yeah [01:08:01] :p [01:08:11] hahaha, holy shit. [01:08:17] perfect, no notes. [01:08:54] [[Tech]]; Quiddity (WMF); /* Upcoming event */ templatize email @; https://meta.wikimedia.org/w/index.php?diff=24466603&oldid=24460918&rcid=26206661 [01:09:03] on that uplifting note, i am going to go get a beer. have a good $time_of_day y'all. [01:09:17] see ya brennen [01:09:21] o/ [01:09:26] o/ [01:10:17] it's also probably time for me to make dinner. Have a lovely internet time, folks. [01:10:25] Alas no beer for me, I've had my allotted midflight drink [01:10:37] enjoy thcipriani ^^ [12:36:01] [[Tech]]; Elitre (WMF); /* Upcoming event */; https://meta.wikimedia.org/w/index.php?diff=24469384&oldid=24466603&rcid=26210918 [15:32:52] Reedy: Umm, what's with the emails you just sent to list? [15:33:02] I didn't [15:33:17] Someone else just mentioned them [15:33:28] https://lists.wikimedia.org/hyperkitty/list/wikitech-l@lists.wikimedia.org/thread/KZ2GKPUW4SJQKDSJVIHEHSABQ3RC3I3T/ [15:33:55] Hmm, lets see, what's the arc header [15:34:35] X-Receiver: dao@it-world.at [15:35:23] SPF: sender does not match SPF record (softfail) [15:35:36] lol [15:36:06] Does mailman not support ARC? [15:37:54] (2.2 points, 4.0 required) <-- interesting that an SPF softfail + invalid DKIM is not enough to make something spam according to mailman [15:39:56] DKIM causes weird mailing list issues IIRC [15:40:48] I think its more on send side than the recieve side though [15:41:27] Usually the main issue is the footer that mailing lists add to the mail invalidates the signature [15:42:34] and maybe whatever headers they change [15:46:08] TheresNoTime: btw, i was thinking of what you were talking about last night with CR office hours [15:46:41] Maybe a page on wiki where people are allowed to list 1 patch they are having trouble getting reviewed would be an interesting idea to explore [15:47:02] A sort of "This is my current most important patch" [15:47:17] hi ppl, somebody can tell me which of these monitor is better? dell e1920h or Aoc E970swn [15:47:20] ? [15:47:38] Guest61: No? [15:47:54] This isn't a channel about monitor hardware [15:48:12] you re right [15:48:16] srry about that [15:48:30] Guest61: no worries [17:47:37] TheresNoTime: So i wonder what you think of https://www.mediawiki.org/wiki/Code_review/patch_board . The basic idea, to prevent things from getting stuck, a board where people can nominate the patch they have that they most wish was reviewed [17:47:45] I'm really not sure if this is a good idea or not [17:47:54] but i figure it can't hurt to try it out [19:30:31] bawolff: worth trying, thank you! [19:30:57] I'm jetlagged to heck, apologies for not replying earlier