[listenbrainz-android] 14dependabot[bot] opened pull request #281 (03dev…dependabot/gradle/dev/org.jetbrains.kotlin-kotlin-gradle-plugin-1.9.20): Bump org.jetbrains.kotlin:kotlin-gradle-plugin from 1.9.10 to 1.9.20 https://github.com/metabrainz/listenbrainz-androi…
2023-11-15 31916, 2023
BrainzGit
[listenbrainz-android] 14dependabot[bot] opened pull request #282 (03dev…dependabot/gradle/dev/com.google.devtools.ksp-1.9.20-1.0.14): Bump com.google.devtools.ksp from 1.9.10-1.0.13 to 1.9.20-1.0.14 https://github.com/metabrainz/listenbrainz-androi…
2023-11-15 31912, 2023
BrainzGit
[listenbrainz-android] 14dependabot[bot] opened pull request #283 (03dev…dependabot/github_actions/dev/actions/github-script-7): Bump actions/github-script from 6 to 7 https://github.com/metabrainz/listenbrainz-androi…
2023-11-15 31903, 2023
lusciouslover joined the channel
2023-11-15 31945, 2023
lusciouslover has quit
2023-11-15 31934, 2023
lusciouslover joined the channel
2023-11-15 31936, 2023
lusciouslover has quit
2023-11-15 31928, 2023
bitmap
going to deploy some changes to the haproxy "timeout client" parameter on the main pg cluster to try and resolve some sitemap queries being prematurely killed...
aerozol: yes, I do. but I'll go after the OSS projects first. plex has a budget for this shit.
2023-11-15 31949, 2023
Rotab has quit
2023-11-15 31912, 2023
Rotab joined the channel
2023-11-15 31934, 2023
d4rkie joined the channel
2023-11-15 31924, 2023
texke has quit
2023-11-15 31936, 2023
texke joined the channel
2023-11-15 31951, 2023
monkey
Moin moin
2023-11-15 31937, 2023
yvanzo
O’Moin
2023-11-15 31932, 2023
monkey
moineau
2023-11-15 31943, 2023
monkey
aerozol: Could you please share edit/commenting access to your doc with the MeB team? I'd like to leave some comments
2023-11-15 31955, 2023
monkey
Or to ansh and myself I guess
2023-11-15 31909, 2023
mayhem
moin moin!
2023-11-15 31912, 2023
mayhem
Etihad's wifi is the first wifi that isn't redonkulously expensive *and* it works. I'm impressed.
2023-11-15 31902, 2023
aerozol
monkey: oops, done
2023-11-15 31920, 2023
monkey
Fanks
2023-11-15 31928, 2023
zas
I made a change to LE auto-generated certificates, to fix an issue making them incompatible with OpenSSL 1.0.2 (which is still in use, especially by some versions of Picard linked to Qt5). It breaks very old Android devices though (more than 6 years old, running <= Android 7.1). This is a tradeoff, but we were provided an expired root cert (see https://www.openssl.org/blog/blog/2021/09/13/Lets…)
It will apply on cert renewal (somewhen between now and 3 months), but I forced the renewal for coverartarchive.org (and its beta)
2023-11-15 31944, 2023
zas
the whole fix is a bit tricky, since it implies installing a recent version of dehydrated, a specific config for it, and an unexpected dependency (hexdump)
[listenbrainz-server] 14amCap1712 opened pull request #2629 (03deduplicate-listens…deduplicate-msids-2): Resolve duplicate MSIDs when fetching MSID data https://github.com/metabrainz/listenbrainz-server…
2023-11-15 31926, 2023
d4rkie has quit
2023-11-15 31959, 2023
d4rkie joined the channel
2023-11-15 31959, 2023
dseomn has quit
2023-11-15 31919, 2023
dseomn joined the channel
2023-11-15 31941, 2023
mayhem
lucifer: since, we recently learned the lesson the hard way, are you planning on opening a PR that updates the backup of the tables that were recently changed?
2023-11-15 31908, 2023
lucifer
mayhem: which recently changed tables?
2023-11-15 31901, 2023
mayhem
the last set of three PRs.
2023-11-15 31913, 2023
mayhem
> Utility tool for migrating listens to new schema table.
2023-11-15 31900, 2023
mayhem
there are new messybrainz tables that have not been added to backups yet. (tbh, I have not loaded the 3rd PR yet)
2023-11-15 31917, 2023
lucifer
i see, those are only there temporarily.
2023-11-15 31941, 2023
mayhem
even messybrainz_redirect?
2023-11-15 31903, 2023
lucifer
yup opening more PRs to fixup the remaining tables, we shouldn't need the redirects then.
2023-11-15 31925, 2023
mayhem
ok, good. just making sure we're not going to miss this again.
2023-11-15 31933, 2023
lucifer
replacing duplicate msids everywhere else in LB.
2023-11-15 31915, 2023
lucifer
yes makes sense. if after cleanups we still feel the need, i'll add it to backups.
2023-11-15 31926, 2023
lucifer
opening a ticket so that we don't forget
2023-11-15 31912, 2023
Sophist-UK has quit
2023-11-15 31901, 2023
outsidecontext
lucifer, mayhem: I'm trying to figure out how I can determine via LB API the oldest listen for a user. If I choose a min_ts too low I get no listens in the listens output. I need to find one that is reasonable close to the oldest listen, but I don't know which this is. Only solution I found right now was to search backwards through the listen history, but that's less the ideal. Any ideas?
outsidecontext: as far as I know there's no way to get that info via the api currently.
2023-11-15 31906, 2023
monkey
However we do load it for the user web page, so I don't think it would be an issue to add `oldest_listen_ts` to the API output. I don't know if that would imply a speed penalty though, but they will know.
2023-11-15 31928, 2023
outsidecontext
Somethingime this would be cool, if it is feasible without much impact
2023-11-15 31948, 2023
monkey
I think it might be a two line change, I'm looking at it now.
2023-11-15 31904, 2023
monkey
Yep, should be super easy and have no impact 🎉
2023-11-15 31939, 2023
outsidecontext
Otherwise the fastest workaround I found was querying the activity history for all time, search for the oldest year with listens, then search backwards from the end of that year
2023-11-15 31917, 2023
outsidecontext
Still involves quite some requests
2023-11-15 31926, 2023
monkey
Yeah, way too much work for simple information
2023-11-15 31904, 2023
outsidecontext
Sound cool, thanks monkey
2023-11-15 31909, 2023
BrainzGit
[listenbrainz-server] 14MonkeyDo opened pull request #2630 (03master…oldest-timestamp-api): API: Add oldest listen timestamp to /user/<username>/listens API response https://github.com/metabrainz/listenbrainz-server…
2023-11-15 31923, 2023
monkey
outsidecontext: ^
2023-11-15 31922, 2023
kellnerd
> Add tests for oldest_listen_ts in /lsitens payload
To avoid dropping then readding the relevant strings
2023-11-15 31955, 2023
reosarevok
(it moves a bunch of stuff out of /admin
2023-11-15 31956, 2023
reosarevok
)
2023-11-15 31905, 2023
reosarevok
I'm rebasing that now
2023-11-15 31909, 2023
yvanzo
thanks
2023-11-15 31916, 2023
bitmap
regarding the sitemaps, I fixed the the overall sitemaps last night by increasing an haproxy timeout which was killing some queries. the incremental sitemaps (which was what the ticket is about) are still broken
2023-11-15 31913, 2023
bitmap
I believe the issue is that they were down for too long, and are now finding too many changes; normally there would not be 50,000 changes to RG alias pages in a single hour, so the code doesn't account for it.
2023-11-15 31928, 2023
bitmap
I'm just going to reset the *_replication_sequence columns in sitemaps.control so that it doesn't process all these old packets
2023-11-15 31940, 2023
yvanzo
👍
2023-11-15 31953, 2023
wargreen has quit
2023-11-15 31918, 2023
wargreen joined the channel
2023-11-15 31905, 2023
bitmap grumbles at this po2json warning
2023-11-15 31914, 2023
reosarevok
Hmm
2023-11-15 31938, 2023
reosarevok
yvanzo: are we not translating pages for admins, or pages for anything that requires privileges?
2023-11-15 31932, 2023
reosarevok
I see you're adding l_admin to things like "Transclude WikiDocs" which is not really "admin" as such
2023-11-15 31955, 2023
yvanzo
reosarevok: admin, not just account admin, but I didn’t include area/inst editors.
2023-11-15 31905, 2023
yvanzo
so it includes transclusion editors indeed
2023-11-15 31937, 2023
yvanzo
and also relationship editors and banner message editors
