#metabrainz

morning

ruaok: ^ that's me uploading a 15gb file to google docs... :)

not great to hear that the cable didn't work for you. crappy ISP router?

ha, nice, alastairp

I've seen 600 mb from my router, so that is not likely it. I should move my router back with the cable I had it on and check to make sure it gets a gig net connection.

alastairp: nudge nudge, please review mbid-mapping soon. or let me merge it and then do your typical post mortem review...

ruaok: yep, planning those two reviews for this afternoon

thx

I'm finishing out the last board bit of business for the year and then I'll have a look at your work on the playlist PR.

Mr_Monkey: ping

pong

I'm reading your playlist.py view code.

that one code block looks quite familiar. :)

but it raised one issue with me -- when I wrote that block I missed an important part..

namely if a playlist is public, but the person who loaded the page is not the owner of the playlist, they should not be able to edit the playlist. are you taking care of that in the front end?

(it is taken care of by the backend, but it would be better to now allow the user to attempt it if they are not allowed to do it)

Users shouldn't see any buttons to edit or delete a playlist that is not theirs, but that could always be bypassable, let me see if I do any other check

perfect. that is all we need. the API will be the final enforcer.

So indeed a user would need to be very crafty to show edit/delete UI if they are not the owner, but I'll also add some more checks if I'm missing any before API calls.

OK ruaok, that's pushed.

kewl

zas: when you're around I can try re-enabling the archive_command and seeing if's unstuck

I'm around

ok, let's see

2020-12-07 15:19:53.197 GMT @ 92 LOG: parameter "archive_command" changed to "barman-wal-archive barman $HOSTNAME %p"

:(

LOG: archive command failed with exit code 120

damn

can you increase log verbosity somehow?

I'm not sure if it's possible, but maybe a wrapper script that invokes barman-wal-archive can capture more details

when I had it set to /bin/true barman actually reported continuous archiving: OK, so maybe it will allow creating a new backup in that case

I can try /bin/true again, create backup, then switch back

ok, try that

backup started ok

backup finished, but archive_command still fails

zas: I believe it was an environment issue after all: $HOSTNAME was not in env. I hardcoded it to 'floyd', and don't see any issues

but it doesn't explain why $HOSTNAME is no longer in env or the 120 exit code, which I couldn't reproduce

alastairp : There are failing tests on LB in all the open PRs, unrelated to the PRs themselves.

Tests started failing from this PR onwards: https://travis-ci.org/github/metabrainz/listenb...

Is that the phantom failing test you've been struggling with and hence something you're aware of? If so, sorry to slowpoke…

hi

that's not the specific test that I was trying to fix, but it's one of the failures that I was seeting

seeing

sigh. what a day

Freso: can I request to go early in the list, I have to run to the post office afterwards, and it's only open for another hour or so

OK, thanks

It’s International Civil Aviation Monday!

No mailed in reviews, and alastairp requested to go early, so…

alastairp: Go!

(Others up for reviews: bitmap, reosarevok, Freso, Mr_Monkey, ruaok, yvanzo, zas, shivam-kapila, jmp_music, CatQuest – anyone else who wish to give review, please let me know ASAP!)

hi there!

:D

hi alastairp!

last week I got listenbrainz tests working on jenkins, to avoid the slowdowns on travis

I reviewed some CB pull requests

and had a quick look at some playlist PRs that ruaok opened, but am behind on them. Planning on finishing that this evening

we ran into some issues in LB with a new version of pip being released, and breaking some things. I spent some time trying to unwind the interaction between pip, brainzutils, and listenbrainz. almost got that cleaned up

this coming week will probably be quieter, because tomorrow is a holiday and today is the day between the weekend and holiday that everyone takes off anyway

bitmap: next

hey

last week I finished working on/updating PRs for MBS-11092 and the improvements to editor JSON handling

MBS-11092: Some issue with userscripts/plugins and beta site https://tickets.metabrainz.org/browse/MBS-11092

those were deployed to beta and should all get released today

did a bunch of code review, caught up on most recent MBS PRs I think

made some fixes to our postgres setup, firstly making sure the services come back up if the containers are restarted (which happened on accident recently), and updating syswiki to document how to bring those up

yesterday I was helping debug issues with the postgres WAL archive_command, and important part of our backup system, which broke due to a missing environment variable

it seems to be resolved as of today, but it's unclear why the environment changed. I guess it happened when the containers restarted

that's mostly it I think, so fin! reosarevok go

Hi!

I spent some time on a couple tickets that had a lot of votes but hadn't been looked at recently: MBS-8371 / MBS-4548

MBS-8371: Search in edit notes https://tickets.metabrainz.org/browse/MBS-8371

MBS-4548: Aliases/Edit artist credit page has no link to see what uses the artist credits https://tickets.metabrainz.org/browse/MBS-4548

Hopefully the first work on the main server and doesn't start always timing out, but I guess we'll see :)

sounds like something we might be able to add indexes for

Other than that, smaller fixes, including MBS-11264 which had annoyed me before but I was too lazy to deal with until someone else reported it :p

MBS-11264: Guess Case > French mode > chain of c combined with ' incorrectly convert it in uppercase https://tickets.metabrainz.org/browse/MBS-11264

And added a few more instrument images since that code is in beta now

CatQuest: you? :)

Meow! `😺`

This week I've stood on my head getting ready for later this month :D

Weather continues to be dismal and my white phlox plant died :(

on the upside I've finished up some book and music release backlog!

aaand thanks to bitmap and reosarevok we now have IROMBOOK images!!

👏 👏 👏 https://beta.musicbrainz.org/instrument/9290b2c... 👏 👏 👏

aaand lastly 2 days from now it is my birthday :DD

⛎

fin, go freso!

o/

Not done much to report on. :/

fin.

Mr_Monkey: Go!

Hello !

Last week I worked on fixing BB-544 and took the oportunity to revamp the editor profile form

BB-544: Unable to change my BB Display Name to match MB account name https://tickets.metabrainz.org/browse/BB-544

Just finished that today

I also continued work on the Playlists feature for LB

I think we're *actually* there after today

lso did a bit of research in edition formats for BB

lots of little bits to finish still, but hey..

Worked on various upkeep tasks

and kept an eye on the forums

And finally some PR reviews :)

That's it for me, zas go !

pick someone else.

ruaok: Go!

heh

last week I spent a lot of time setting up test.lb.org and fixing bugs.

I finished dealing the data leak fallout and reporting the breach in california.

minor bits of meb work here and there.

then on friday I finished off some troi work, including creating a troi-bot that will soon create recommendation playlists for other users on LB.

that might get put in place as early as this week. 🤞

fin. zas if he responds in 30 seconds, otherwise yvanzo

*drumroll*

Looks like yvanzo :)

still a few seconds…