yeah, it's unclear exactly what happened but the ProcessSubscriptions script was hanging for almost 20 hours, and the next daily script didn't start because the previous one was still running
the daily script for today just started running so you should get a subscription email summarizing the previous three days soon
MyNetAz has quit
MyNetAz joined the channel
lusciouslover joined the channel
minimal has quit
GautamShorewala[
<aerozol[m]> "If it’s *optional*, I don’t..." <- But there should be some way to have users play the music or are we planning to implement something else for that
aerozol[m] has quit
aerozol[m] joined the channel
jasje[m]
<GautamShorewala[> "But there should be some way..." <- Gautam Shorewala: songs already play through youtube music and spotify, wym?
Kladky joined the channel
outsidecontext[m has quit
lucifer[m] joined the channel
lucifer[m]
bitmap: fine to remove the aretha database
<fettuccinae[m]> "mayhem: is it viable to re-..." <- the code to generate YIM stats has changed since last year, the frontend for previous years would need to be updated to handle new data format and it could work then. but even then it would work only for 2023. for 2022 and 2021, it would be way too much effort.
Would like to get that deployed along with your playlist cover work, otherwise it sends you straight to the bottom of the page and we can't enjoy the image :p
q3lont joined the channel
q3lont has quit
ansh[m]
<monkey[m]> "Also ansh do you have a minute..." <- Sure! I'll review it
ansh: Just noticed something: the playlist description isn't visible, something going on with the `.wipedia-extract .content` class. Happy to look into it if you want, let me know.
monkey[m]: Correct, that means it's also broken on the artists page.
monkey[m]
Yeah, I think we never noticed because either we get a full wiki paragraph or we get nothing, but we probably never get a very very short wikipedia blurb
holycow23[m] joined the channel
holycow23[m]
Hi, I was planning to work on [LB-1737](https://tickets.metabrainz.org/projects/LB/issues/LB-1737?filter=allopenissues), had some questions regarding where exactly would the Chart be displayed and does it follow the same 'duration' based mechanism?
BrainzBot
LB-1737: Create top artists graph showing album details
holycow23[m]
* Hi, I was planning to work on [LB-1737](https://tickets.metabrainz.org/projects/LB/issues/LB-1737?filter=allopenissues), had some questions regarding where exactly would the Chart be displayed and does it follow the same `duration` based mechanism as we can see on the `stats` page?
BrainzBot
LB-1737: Create top artists graph showing album details
ansh[m]
<holycow23[m]> "Hi, I was planning to work on [..." <- It should likely be displayed on the stats page itself. Yes, it should adhere to the same range options we currently have.
that is the only known stats issue. but there was another strange bug mentioned to support@ : "strange auto link behaviour".
unsure what is going on there, please take a look when you can
lucifer[m]
i think its working fine at least right now.
looking at the user's current charts.
yes that email i saw and am looking into it already.
monkey[m]
I was looking into the artist origin map as well just now, does indeed look fine at the moment, so maybe it's just lagging behind other stats and making it look broken
holycow23[m]
I was working with the similar methodology, and was able to generate the below stats, do you think I could formalize this into a PR and assign the ticket to myself?
holycow23[m] uploaded an image: (23KiB) < https://matrix.chatbrainz.org/_matrix/media/v3/download/matrix.org/iHmYaijFljUXPRSDgnobsUyn/image.png >
lucifer[m]
mayhem, monkey: okay i think i know why artist origins lags sometimes. i'll move the stat over to spark and that should fix it.
monkey[m]
Ooohhh! What is the issue as you see it?
lucifer[m]
or at least make it consistent in lagging with other stats
we used to recompute all stats daily and create a new database for each stat daily. now that is done once a week and rest of the days incremental stats only updated records for user who listened to listens in that range. artist maps calculation checks if a row for the stat already exists in the db and if so it serves that instead of recomputing it from scratch.
with a fresh db daily, artist map stats were regenerated daily. but now that db are generated weekly, rows already exist in the db and the artist map data is not created again.
it must have fixed itself earlier today because a new database was created for the stats today. it will go stale again tomorrow.
mayhem[m]
want me to respond on the forum?
lucifer[m]
sure thanks!
reosarevok[m]
Releasing prod
julian45[m] joined the channel
TOPIC: MetaBrainz Community and Development channel | MusicBrainz non-development: #musicbrainz | BookBrainz: #bookbrainz | Channel is logged and not empty as it is bridged to IRC; see https://musicbrainz.org/doc/ChatBrainz for details | Agenda: Reviews, Summit dates (mayhem), staff/admin SSO (julian45/zas)
holycow23[m]
* I was working with the similar methodology, and was able to generate the below stats, do you think I could formalize this into a PR and assign the ticket to myself monkey
have you seen an issue such as this before? I can trigger a new dump if you'd like
monkey[m]
<holycow23[m]> "I was working with the similar..." <- holycow23: Yes, by all means you can assign the ticket to yourself if you are interested in working on it! The mockup looks good, that's more or less what I had in mind. With a little bit of interactivity to show album names on hover it'll be really nice
lucifer, mayhem: I have one of those phantom bugs where everything works fine when run locally, but the CO tests stall indefinitely. They pass consistently locally and fail consistently in CI, presumably at the same point looking at the logs.
There is no failing test and no useful output to help figure out the issue, just hanging tests. Do you have a special way to diagnose these issues? Some way to trigger tests with verbose output for example?
Actually lucifer I see that a bunch of recent commits merged from PRs with passing CI have a stuck spark test suite that gets canceled after some time. Could be related to recent incremental stats changes?
monkey: yes, the spark test suite is broken atm and can be ignored.
Jade[m] joined the channel
Jade[m]
<mayhem[m]> "I'm hoping that eventually we..." <- Yeah those templates support HTML in the strings, it's just wiring them up into the perl. I ran into a roadblock getting some of the more complex tests to work with them
reosarevok[m]
yvanzo: I see there's a few extra changes on the docker side this time, do you want to draft the release?
Can a French speaker (zas?) take a look at MBS-13927 ? The user seems quite confused so the edits might be all wrong too for all I know
Jade[m] uploaded an image: I've just written a new endpoint that allows sending in the MJML, which it then renders to HTML + text and emails. It doesn't have access to the i18n stuff I did, as that happens in the template itself, but it has all the rest of it (34KiB) < https://matrix.chatbrainz.org/_matrix/media/v3/download/ellis.link/Yj5EY9sNwnC3jUtY19vrNS4wKgoeOayk >
Jade[m]
The bonus is you can keep the templates in the repo with the project and you don't need to recompile/redeploy the service to update templates in that case
Once CI finishes it'll be availible at metabrainz/mb-mail-service:main if you want to experiment
LMK if a bulk endpoint or a preview endpoint like the ones for the templates would be useful
bitmap[m]
zas: hey, can you make it so that emails to npm at metabrainz dot org are forwarded to me?
zas[m]
The email address already exists?
bitmap[m]
yes, according to syswiki only akshaaatt has access to it currently
<holycow23[m]> "Sure, yes I have added that..." <- Does this look good?
holycow23[m] uploaded an image: (34KiB) < https://matrix.chatbrainz.org/_matrix/media/v3/download/matrix.org/qMQWqROoaWleqNVkmOeILmzr/image.png >
minimal joined the channel
lusciouslover joined the channel
dvzrv joined the channel
BrainzGit
[musicbrainz-server] 14mwiencek opened pull request #3470 (03production…xgettext-js-pkg): Install xgettext-js from org-scoped package rather than git https://github.com/metabrainz/musicbrainz-serve...
dvzrv has quit
yvanzo[m] joined the channel
yvanzo[m]
bitmap: Thanks for the pointer. I will handle it.
reosarevok: No thank you.
reosarevok[m]
yvanzo: Ok, but I see there's more there than the usual update, so the tag description shouldn't just be "Upgrade MusicBrainz Server.". Should it just be "Set up SIR development via Docker Compose environment variables and upgrade MusicBrainz Server." then?
Anyway, post meeting is fine
<BANG>
Hi, and welcome to another MetaBrainz Monday Meeting!
Jigen
:D
reosarevok[m]
We have two mailed in reviews today!
derat says:
"I finally dug into MBS-13920 and sent two small PRs that I think will fix it. I also started working on a few URL batch edit tickets (MBBE-91, MBBE-92, MBBE-93)."
