for instance i am trying to view the headers/body of a message sent earlier today but get this error.
2022-03-09 06846, 2022
lucifer
ok it appears exim does not retain sent email :/
2022-03-09 06811, 2022
akshaaatt
Wow aerozol! Love the additions you've made! 👌👌👌
2022-03-09 06835, 2022
akshaaatt
aerozol the modern MB look looks fire to me and I think we should be keen to have it that way 👌
2022-03-09 06813, 2022
aerozol
I would expect whatever styles we settle on to get used, more or less, in future MetaBrainz develpments, so it's good to make sure
2022-03-09 06818, 2022
aerozol
Including LB etc
2022-03-09 06857, 2022
akshaaatt
aerozol I think LB would be flexible with any design changes you might have. We even need to better work on some CSS for LB rn.
2022-03-09 06846, 2022
zas
lucifer: exim doesn't retain emails but if they are still in queues. I guess that's exim relay for lb running as container, so docker logs is your friend.
2022-03-09 06841, 2022
zas
lucifer: see my PM about this
2022-03-09 06837, 2022
aerozol
Great! LB has some square buttons but also some very rounded elements in-page so I'm trying to find some solid ground
2022-03-09 06830, 2022
texke joined the channel
2022-03-09 06808, 2022
gayangi81 joined the channel
2022-03-09 06840, 2022
zas
yvanzo, atj: I don't understand the use of sudo -A in sudo -A rsync -avh --progress --rsync-path='sudo -A rsync' user@remote:/remote_path/ /local_path command, if you want to preserve env, use -E, but -A is to ask pass, which doesn't look needed. Now I'm not sure to understand what you are trying to do, what do you try to copy, from where to where?
2022-03-09 06815, 2022
zas
atj: SSH_AUTH_SOCK to env_keep -> not sure if this is needed if one uses sudo -E: try sudo -E printenv | grep SSH
2022-03-09 06847, 2022
yvanzo
zas: Thanks, -E is what I intended to use actually ^^
2022-03-09 06806, 2022
yvanzo
zas: Copied the docker volumes musicbrainz-sitemaps and musicbrainz-sitemaps-logs from yehudi to rakim on yesterday.
2022-03-09 06822, 2022
yvanzo
O’Moin
2022-03-09 06858, 2022
gayangi54 joined the channel
2022-03-09 06847, 2022
yvanzo
zas: Are port numbers in PORT_LIST supposed to be unique? I guess that sharing the same number for both musicbrainz-redis-cache and musicbrainz-redis-cache-test doesn’t clash because these containres are not running on the same host. Should those be differentiated instead?
2022-03-09 06852, 2022
gayangi81 has quit
2022-03-09 06821, 2022
zas
they are supposed to be unique to avoid clashes when moving services around, but they don't have to, only the key have to be unique (it works like a dict)
2022-03-09 06804, 2022
yvanzo
Moved the container musicbrainz-sitemaps (and its redis instance) to rakim yesterday.
2022-03-09 06850, 2022
yvanzo
Planning to move musicbrainz-redis-cache today.
2022-03-09 06836, 2022
yvanzo
Moving musicbrainz-redis-store will require a downtime, will plan with bitmap and reosarevok later today.
2022-03-09 06852, 2022
zas
ok, perfect, tell me when all is done
2022-03-09 06838, 2022
zas
atj: what's with shorewall & IPv6?
2022-03-09 06842, 2022
atj
zas: Shorewall is IPv4 only, there is a separate package called Shorewall6 that handles IPv6
2022-03-09 06804, 2022
atj
they share a lot of code, and the configuration is almost identical
2022-03-09 06822, 2022
atj
yvanzo: I can confirm that "sudo -E" works, did you try that before I made the env_keep change?
2022-03-09 06845, 2022
zas
atj: I guess we need it to be deployed then
2022-03-09 06802, 2022
yvanzo
atj: No I just didn’t type the correct option.
2022-03-09 06810, 2022
atj
yvanzo: ok, zas solved it then :) I've removed the env_keep option
2022-03-09 06823, 2022
yvanzo
atj: yes, thank for your help :)
2022-03-09 06830, 2022
atj
no worries. I clearly need to brush on my sudo knowledge!
lucifer: ^^ I think we need to delay how long before we send the mail about a user's spotify connection being disabled. it seems that a lot of people are not re-connecting. And it makes sense, just because spotify wobbles, our users shouldn't have to do anything.
2022-03-09 06847, 2022
lucifer
alastairp: more fun `timescaledb-2-2.6.0-postgresql-13` is present in the repo but 2.2.0 is not :/ for ubuntu.
was this a spotify error or did LB handle the downtime badly?
2022-03-09 06817, 2022
mayhem
spotify was down.
2022-03-09 06835, 2022
lucifer
spotify had a downtime and the error reported by spotify was probably misconstrued as the user having revoked access to LB.
2022-03-09 06836, 2022
mayhem
I dont think any of this was our problem. (but dont quote me on that)
2022-03-09 06830, 2022
gayangi54 has quit
2022-03-09 06830, 2022
atj
so the user then tried to grant access again to LB, which actually revoked access, but Spotify was down so it didn't work anyway?
2022-03-09 06816, 2022
lucifer
when we detect access revocation eerror, we delete the access tokens on our side and mail the user to reconnect again.
2022-03-09 06812, 2022
lucifer
but spotify was down so yeah it didn't work and we cannot reconnect those users ourself without them reapproving spotify side because we deleted the token
2022-03-09 06813, 2022
mayhem
I wonder if there is any way to retry after the revocation error... because it may have been erroneous/
because maybe we should have some backoff -- and only ever give up after X hours.
2022-03-09 06838, 2022
alastairp
hi
2022-03-09 06840, 2022
texke joined the channel
2022-03-09 06845, 2022
mayhem
essentially allowing for this case.
2022-03-09 06850, 2022
alastairp
lucifer: hmm, is that a specific apt package?
2022-03-09 06803, 2022
lucifer
yes that makes sense. we don't have a delay yet.
2022-03-09 06820, 2022
mayhem
lucifer: lets pick this up tomorrow after the TS upgrade.
2022-03-09 06851, 2022
alastairp
lucifer: I installed `timescaledb-2-postgresql-13` and it came with `/usr/lib/postgresql/13/lib/timescaledb-2.2.0.so`
2022-03-09 06859, 2022
lucifer
alastairp: yes. TS repo has apt package for each TS and PG version combination.
2022-03-09 06841, 2022
lucifer
oh let me try that. i was looking for timescaledb-2-2.2.0-postgresql-11 because the other one is timescaledb-2-2.6.0-postgresql-13.
2022-03-09 06847, 2022
lucifer
mayhem: 👍
2022-03-09 06805, 2022
alastairp
timescaledb-2-postgresql-13 and timescaledb-2-postgresql-11 installed fine for me, though I didn't try to start it up with a database
2022-03-09 06854, 2022
riksucks
monkey: thats a good idea, since, if we do further changes, then whoever is doing changes need not worry about putting Nullish coalescing operator all the time. Is it okay if I implement it in the PR which would have the implementation of hiding?
2022-03-09 06800, 2022
monkey
riksucks: Sure, that'll work
2022-03-09 06806, 2022
texke has quit
2022-03-09 06817, 2022
texke joined the channel
2022-03-09 06834, 2022
alastairp
CB released, gravatars removed
2022-03-09 06838, 2022
reosarevok
Yay
2022-03-09 06802, 2022
alastairp
monkey's review of the uuid cccccc-cccccccc-cccc-cccccccc is gone too!
2022-03-09 06816, 2022
monkey
Nooooo! Not cccccc-cccccccc-cccc-cccccccc !
2022-03-09 06844, 2022
alastairp
it was a good release, huh?
2022-03-09 06835, 2022
reosarevok
I c what you did there
2022-03-09 06847, 2022
alastairp
oh, it was unintentional!
2022-03-09 06848, 2022
riksucks
also monkey, I haven't thought much on the exact implementation (have recently moved into a dorm), so I am going to say what I have in mind, please refine wherever you feel like. My plan is to store keys of the events that I am gonna hide in localStorage. Key would probably be something like `${event.event_type}-${event.timestamp}` (since not all events have id available, like listens).
2022-03-09 06805, 2022
alastairp
hmmm, poking around the CB DB I found a _lot_ of spam comments, just looking into them now to see if they're public or not
OT: GSoc Mentors have been invited. if you didn't get an invite but were expecting one, ping me.
2022-03-09 06822, 2022
lucifer
reosarevok: yeah this has been reported many times before but we have been unable to reproduce it reliably.
2022-03-09 06826, 2022
reosarevok
Oh, ok
2022-03-09 06810, 2022
lucifer
alastairp: mayhem: how should we backup timescale docker volume?
2022-03-09 06817, 2022
alastairp
lucifer: `cp`
2022-03-09 06821, 2022
monkey
riksucks: Yes, I think something along those lines would work. I've come to the same conclusion for deleting listens where we have a similar issue (we recently moved to a mechanism that marks a listen to be deleted, with listens being deleted in batch every hour)
2022-03-09 06838, 2022
mayhem
oh. I have a couple of tables that I could get rid of.
2022-03-09 06840, 2022
mayhem
hang on.
2022-03-09 06850, 2022
lucifer
oh, that seems simple.
2022-03-09 06800, 2022
mayhem
canonical_recordings for instance. that will make the backup easier.
monkey: can we also employ a similar algo to delete keys of listens from localStorage, will prevent stale storage
2022-03-09 06801, 2022
monkey
Yeah, i was just thinking about how to do that. Any ideas?
2022-03-09 06846, 2022
riksucks
an algorithm that checks the oldest timestamp available in the feed, then deletes any event that's in the localstorage that's older than that
2022-03-09 06843, 2022
riksucks
we should probably store these keys in an array, so that the time series aspect is retained
2022-03-09 06853, 2022
alastairp
mayhem: hah, amazing
2022-03-09 06800, 2022
alastairp
!m reosarevok and others working on that ticket
2022-03-09 06800, 2022
BrainzBot
You're doing good work, reosarevok and others working on that ticket!
2022-03-09 06822, 2022
reosarevok
Thanks, I guess
2022-03-09 06824, 2022
monkey
"oldest timestamp available in the feed" does that take pagination into account?
2022-03-09 06833, 2022
riksucks
no, only the first page, because I thought the hide should only work for the first page? that's the assumption I had in mind, because you would want to hide something that is appearing multiple times in your whole UX
2022-03-09 06804, 2022
riksucks
but if something is in seocnd page, it's kinda already hidden. that is how I had in my mind
