#musicbrainz-devel

Project critiquebrainz build #116: FAILURE in 2 min 10 sec: http://ci.musicbrainz.org/job/critiquebrainz/116/

* dufferzafar0: Modified the review entity view to allow viewing revisions

* dufferzafar0: Add 'View older revisions' link to review entity page

* dufferzafar0: Fetch only the required revision in the review.entity view

* dufferzafar0: Fetch only the required revisions in the review.revisions view

* dufferzafar0: Display older revisions link in all cases

dufferzafar: ^ forgot to update RevisionTestCase :)

can you fix please?

Please review/test/comment last iteration of https://github.com/musicbrainz/picard/pull/408

Freso: i may split this PR before merging, one for cover art options revamp, one adding the local files support

:)

https://musicbrainz.org/doc/Development/XML_Web... do we have endpoints for non-release collections?

alastairp: for events at least https://beta.musicbrainz.org/ws/2/collection/13...

I need works

Well, just check with a work collection :)

If it doesn't work, ticket

Given how the events one works I'd expect if it works it'll just be ws/2/collection/id/works

no, it doesn’t work

yes, I’ll make a ticket

Ok! :)

(also meh :( But at least we know now)

but honestly, a little weird that it was “released” without being fully working...

Well, the schema change bits were released during the schema change, I think the rest is still kinda WIP

(not even all the entities can be added to collections at all yet I think? Unsure)

I might very well be mistaken though, I'm not very on top of development right now

sure

OK, added

Thanks!

MBS-8459

I guess I’ll put this task to the side. or use requests on the main website :(

I kinda want to put all these metadata tasks to rest, but we’re coming up against lots of problems

but tickets are getting fixed, so that’s great

I mean, you're our early adopters, there's always going to be some bugs

it has to be perfect all the time, damnit

How do you survive Spain :D

sometimes I retreat to my little english-speaking bubble

Oh, maybe that'd work. Definitely would never expect anything in Spain to be perfect all the time. Or work. Or even start.

Or be there at all?

nah, spain’s here

OK

a pity, because the url on the website uses rowid ;)

That should possibly also be reported. :p

already done

OK, I have a mirror which I can get rowids from

Freso: PR splitted in https://github.com/musicbrainz/picard/pull/411 and https://github.com/musicbrainz/picard/pull/412

pymb people, who has any idea about how we do timeouts? From looking at code it looks like we don’t do anything special, so perhaps defaults are used?

what sort of timeouts?

502s, you mean, or?

sorry, that was unclear

I’m doing some AB stuff. I want to wind the socket timeout down really low, so if the MB request takes too long for whatever reason, we fall back to submitted metadata

I was wondering if we expose the http request timeout as a parameter (we don’t)

but we can use socket.setdefaulttimeout

*however*

in our infinite wisdom we added an auto-retry functionality, which isn’t user-configurable either :(

apparently we import build_opener twice from urllib2, heh

but yeah, I think we basically just used defaults

I wonder if it’s a good idea to expose max_retries

maybe even by a hidden _parameter

that can be changed if you know what you’re doing

like we do with parsers

yeah, I think it probably makes sense to have everything replaceable if needed

ianmcorvidae: re: 8459, I wanted write, but I hacked it by using the website endpoint

so it’s not a requirement for me right at this moment any more

right. I'll do a branch for the easy part that adding the read endpoints is then, and maybe look into the write endpoint stuff later

yeah, good idea