[07:16:59] !log tools.lexeme-forms deployed 2035050d28 (l10n updates: sv) [07:17:02] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.lexeme-forms/SAL [08:09:41] !log adamin downgrading ruby2.5 to previous backport on all buster VMs [08:09:42] dcaro: Unknown project "adamin" [08:10:13] !log admin downgrading ruby2.5 to previous backport on all buster VMs [08:10:16] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [10:06:25] !log admin upgraded ruby2.5 to latest fixed version on all buster VMs [10:06:28] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Admin/SAL [10:29:58] !log paws upgrade jupyterlab T324002 [10:30:02] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Paws/SAL [10:30:02] T324002: Upgrade jupyterlab - https://phabricator.wikimedia.org/T324002 [11:03:52] !log tools.malnadachbot mark for deletion, only maintainer is blocked [11:03:53] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.malnadachbot/SAL [11:09:56] TheresNoTime: hi, regarding T328691 where is the most recent code for linkwatcher? [11:09:57] T328691: [toolsdb] Migrate linkwatcher db to Trove - https://phabricator.wikimedia.org/T328691 [11:16:16] Amir1: that'll be on the VPSs themselves — fairly sure the deployed code and git repo (if there is one..) diverged a long time ago [11:17:17] TheresNoTime: can I haz access then? [11:17:32] Sure thing [11:18:26] Amir1: what's your user/shell name? :D [11:18:39] I think ladsgroup [11:20:22] Amir1: done, it's on `coibot.linkwatcher.eqiad1.wikimedia.cloud` in `/home/beetstra` [11:20:50] have fun with perl :D [11:21:29] (oh and it runs in a screen session because) [11:22:01] * taavi sighs [11:22:39] a mild improvement over `nohup ./the/shell/script.sh &` [11:23:15] 😭 [11:23:22] * TheresNoTime wants to set up `supervisord`, one day™ [11:23:41] why supervisord and not systemd? [11:24:11] honestly? personal preference, and I like `supervisorctl` [11:24:23] (either would be better than a screen session) [11:24:44] obligatory link https://adtmag.com/articles/2014/10/21/~/media/ECG/adtmag/Images/2014/01/systemd.png [11:25:48] how a gif is a png, I don't want to know [11:26:05] D: [11:26:18] thank you for wanting to look at the queries though Amir1 [11:26:40] no promises :D [11:26:41] https://en.wikipedia.org/wiki/APNG [11:27:15] ah interesting [11:27:26] COIBot could really do with a perl dev taking it over (or someone rewriting it... or WMF creating a table which emulates the same sorta thing) [12:04:57] I would vote for a complete rewrite, and I don't think it would be an impossible project, the hardest part is getting it onto a roadmap somewhere. maybe we should've sneakily added it to the Annual Plan :P [12:05:32] could even be an outreachy kind of project? not sure [12:06:11] thanks a lot Amir1 in the meantime for bravely stepping into the Perl wilderness :) [12:09:47] I would like to see some passing investigation into a wmf-backed table of link additions (but I say this with no prior knowledge of the work it would take to do this) [12:13:28] It'll be a lot of work to have a table like that [12:13:56] but I definitely would vote for rewriting the script to python or god forbid rust [12:14:25] Python +1 [12:26:14] oooh it could use eventstreams too.. [14:47:51] @TheresNoTime are you really using screen? tmux is so much better. [14:50:58] *screen my beloved* [14:57:09] * stw is still a screen afficionado [14:57:20] but then again I rarely use it these days anyway [14:57:46] The only thing I regularly run in screen is irssi, but that itself is launched by systemd [14:58:09] (don't question my IRC bouncer setup... it's completely bonkers and I fully accept that) [15:14:49] I never understood people who used either screen or tmux to run production stuff in. [15:15:16] it is a Bad But Easy Idea(tm) [15:15:58] I mostly use it to keep from losing emacs sessions when my internet connection flakes out. I used to use it to keep a single session up while I moved from home to train to office. [15:16:59] These days, I've finally dragged myself into the modern era and do most of my work in VSCode on my laptop, so not so much of an issue. [15:19:28] I think irssi is the only "production" thing I've got running in screen. [15:19:54] and it's only running in screen so I can quickly attach to it if I'm missing my usual IRC client [15:20:10] everything else is either systemd or docker. [15:20:48] All languages, editors, utilities, and libraries are valid if they help someone get something done. ;) [15:27:04] Hi. Anyone from the PAWS team here?  I was wondering if there were any changes to JupyterLab (zoo) recently. It's behaving strangely: every few minutes it slows down my browsers (FF, Chrome want to close it because of "a script"), the view resets itself and jumps to the top of the notebook, all cell numbers disappear, but the kernel stays running. [15:27:04] It also refuses to load recently saved files, unless I duplicate them (same size on the disk) - in which case the copy opens fine... for some time. [15:28:24] hmm. jupyterlab was upgraded. When did you start seeing things? [15:28:37] some 3 hours ago [15:29:07] it's pretty much unusable now :( [15:29:23] Seems consistent. [15:29:26] Let me take a look [15:30:13] but I noticed it being slow last time we spoke, this today is just another level [15:40:30] Do you have a minimal example that I can use to recreate what you're seeing? [15:43:30] You could try to open one of the three notebooks (original, its copy, and its copy). That's under PonoRoboT > HR popisi stanovništva > Py - Popisi stanovništva - enwiki - new... [15:54:11] hmm...In my, naive, view that seems to be working alright [15:54:33] Let me leave it running for a few minutes and see what happens [15:57:21] not sure if you can actually run the cells. this starts happening when I'm approx. half way through the notebook (shift-enter'ing) [15:58:31] and it's still not opening the original file, meaning it's all blank [15:59:03] Mine fails when it gets to a file not found, which in my case isn't there. But it looks alright otherwise [16:00:04] that folder is set to not be public, btw (re @wmtelegram_bot: You could try to open one of the three notebooks (original, its copy, and its copy). That's under PonoRoboT > HR popisi ...) [16:01:48] it's not public because it contains non-free xslx files. but it's been working fine for many months like that [16:03:06] storing non-free content in PAWS is maybe not a great idea... [16:05:04] it's public (census) data, just not CC-... [16:12:53] Ponor: do simple notebooks work for you? `print("foo")` or the like? [16:14:12] they do. even these work until they don't, and after a minute they work again [16:15:28] and because my browsers are complaining I'm thinking there must be a (JupyterLab) script running on my end that's misbehaving [16:19:16] (i did restart my laptop; there were no software updates for 2 weeks at least - so it must be because of the JLab upgrade) [16:20:14] I agree it sounds like a jupyterlab problem. I'm having trouble recreating it to be more helpful [16:22:44] i thought it'd be hard to reproduce.   I'll probably switch to Classic Notebooks, but would like to know if there are any plans to upgrade JLab to the newest version (4?) any time soon? [16:24:22] Indeed, I'm trying that route (upgrading to 4.0.1) as we speak. Seems to have broken something in the extensions, trying to track that down [17:32:16] !log tools.stewardbots ./stewardbots/StewardBot/manage.sh restart # RC not working [17:32:18] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.stewardbots/SAL