#metabrainz

Morning!

Morning!

ruaok, iliekcomputers: Someone wants to interview MeB for a Python podcast - obviously I'm not the right person for it given I have written like two lines of Python for MeB, but if you are interested see email ;)

I think there's some glitch in apple music app for Android, I am using it for free after my membership is over :)

you used credit card for membership?

[bookbrainz-site] prabalsingh24 opened pull request #483 (UserCollection…test-add/remove-items): tests for `collection/add` and `collection/remove` https://github.com/bookbrainz/bookbrainz-site/p...

white_snack: yup, but my membership expired at July 31st :)

It's not working in MacOS as well, only in android

ruaok: ah, right. if this code is completely standalone then I think it does make sense to have a separate image

moooin!

alastairp: not only standalone, but also running on a different machine.

and doesn't need any code from lb or msb?

alastairp: iliekcomputers : do either of you want to take this podcast interview?

alastairp: none, at all.

perfect, keep it separate then

forward me the email?

Mr_Monkey: hi

You mentioned about checking the error message in tests

for tests that are supposed to throw error

Right now i am only checking status code

done.

but the thing is, whenever we catch an error, we do `next(err)` and it does not return a json response

rather it returns and html page with status code 500

I think we should replace all the `next(err)` with this https://github.com/bookbrainz/bookbrainz-site/b...

`error.sendErrorAsJSON`

then `res.body.error`will have the error we're looking for

ruaok: I'm happy to take it.

alastairp how do you feel about it?

iliekcomputers: did you see my questions from last night above? about the same/separate image?

I did not. Let me read the backlog

This is the messybrainz-labs code?

yes

Do you think it'll eventually get merged into LB, given that messybrainz is in LB already?

I've already got it working as a separate image and it would be easy, I think, to make a cron image out of it.

yes, that is the whole point -- how do we do this?

I don't have strong opinions on whether it should be a different image or not. If piggybacking gives us some benefit, we should do it, otherwise standalone sounds good to me too.

However, we should merge the code into LB eventually

ok, great. I'm hoping to have a PR today or tomorrow. its been lingering for too long.

Sounds good.

ishaanshah: https://labs.api.listenbrainz.org/artist-credit...

artist msid lookup is now live on labs.api -- please note that in this case the [artist_credit_mbids] was also renamed to [artist_credit_mbid] for consistency sake.

please use this endpoint from now on.

Cool, I will update it

Thanks :)

now I have a thicket of PRs to prepare.

np!

ruaok: Freso: can we reach out to this user to politely ask them to stop spamming the same artist every 2 minutes 24/7? https://listenbrainz.org/user/hds

1196 listens / day seems a bit much

should we just log into the DB and change their submit key?

eh, i'd like to ping them once as a warning before disabling submissions etc.

also interested in how MusicBrainz handles cases like this, just outright ban or warning first?

when its clearly spamming its delete account. if not clearly spamming, warn.

Freso: over to you then.

[bookbrainz-site] prabalsingh24 opened pull request #484 (UserCollection…test-deleteCollection): Test: Delete Collection https://github.com/bookbrainz/bookbrainz-site/p...

ruaok: Internal server error for the POST endpoint

😭

uh oh

ahhh, the tests I am writing *just* caught this. will have it fixed in a few.

thanks :)

Hi prabal! I don't think we need to return `sendErrorAsJSON`, as I'm sure I've dealt with errors with a regular `next(error)` in the recent past. Let me look for it.

prabal: Here it is:

Where the error is thrown (using `next(err…)`): https://github.com/bookbrainz/bookbrainz-site/b...

And and example of testing for the error message: https://github.com/bookbrainz/bookbrainz-site/b...

See if that method works fo you. I'll have a quick check of your testing code to see if there's any obvious deal breaker

ishaanshah: try again!

ruaok: Working!

fabu. :)

Mr_Monkey: Yeah`resonpse.res.statusMessage` has the error message.

I was checking `response.body`earlier

it has html page

Yeah, that's where I would have looked at it too.

yeah if it were returning json response then it would have been in response.body

I'll make that change

[listenbrainz-server] mayhem opened pull request #1013 (master…labs-api): Add dataset hoster based labs.api.listenbrainz.org https://github.com/metabrainz/listenbrainz-serv...

OK. I'll start having a look at your newly opened PRs

ruaok: the PR looks good to me. Just a few nits, imo the files should be in `listenbrainz/labs/api` instead of `listenbrainz/api/labs` as i could see people thinking `listenbrainz/api` contains the actual API as well. Also, we should pin the dataset-hoster requirement to a git tag.

I'm cool with the tag...

And I see your point on the confusion.... But labs is under the api concept, not the other way around.

hmm

Rename the top ish level dir to labs-api ?

maybe `listenbrainz/webserver/views/labs`

Too hidden, I think.

that works too, but `labs_api` instead, so that it's a valid python name

Yeah, ok.

Will fix after lunch.

thanks!

I'll add a read me too to point out that this is not the main api.

prabal: I just noticed something: in all three open PRs with tests, you use `beforeEach` to set up the loggedInUser and request agent, but it should be `before` instead, since it only needs to be done once.

Oh yeah I'll change that

Same for `afterEach`-> `after`