#metabrainz

[listenbrainz-server] paramsingh merged pull request #590 (production…restart-spotify-imports): Add script to restart spotify imports for users https://github.com/metabrainz/listenbrainz-serv...

ok, i restarted imports for all users (with disabled emails for a few hours, so that we don't send any repeat emails to users with actual non 404 spotify errors)

ferbncode: Tests from my latest commit seem to be failing, but locally they run successfully for me.

spellew: How do you run the tests locally?

You can run them in the same way as jenkins using `docker-compose -f docker/docker-compose.test.yml up --build`

This is the test docker file ^

I'll try that

👍

I was running the file with Python after dockering into the container

okay, did you try to run your local changes locally on CB site?

yvanzo, bitmap : CAA queue is growing see https://stats.metabrainz.org/d/000000075/alerts...

hi gr0uch0mars

thr bug is fixed and internationalization is also complete

hi amCap1712

great, I saw the ratings PR

yeah once it is merged i'll open internationalization pr

good, I'll review as soon as possible

and by the way, I've sent the First Evaluation to Google

In case you receive a notification

the workflow for importing strings is set up. we just need to run a command and all translated strings will be fetched.

good!

i too completed the evaluation

so we are ahead of schedule now

yeap, I think this week marks a great milestone for the project!

yup

🎆

we can't slow down!!! hehehhe

yeah i agree

nevertheless, congrats!!! 👏

congrats to you as well

you're making me think about making an iOS version! hahaha

maybe soon after gsoc ends

bitmap: is there a reason https://tickets.metabrainz.org/browse/MBS-10257 is not like that already?

MBS-10257: List part number on work-work "part of" rels both ways

amCap1712: I'm having problems navigating to an Artist:

```

2019-06-25 10:07:12.395 7089-7180/org.metabrainz.android E/AndroidRuntime: FATAL EXCEPTION: OkHttp Dispatcher

Process: org.metabrainz.android, PID: 7089

java.lang.IllegalStateException: closed

at okio.RealBufferedSource.request(RealBufferedSource.java:70)

at okhttp3.logging.HttpLoggingInterceptor.intercept(HttpLoggingInterceptor.java:251)

at okhttp3.internal.http.RealInterceptorChain.proceed(RealInterceptorChain.java:142)

at okhttp3.internal.http.RealInterceptorChain.proceed(RealInterceptorChain.java:117)

at org.metabrainz.mobile.data.sources.api.HeaderInterceptor.intercept(HeaderInterceptor.java:41)

at okhttp3.internal.http.RealInterceptorChain.proceed(RealInterceptorChain.java:142)

at okhttp3.internal.http.RealInterceptorChain.proceed(RealInterceptorChain.java:117)

at okhttp3.RealCall.getResponseWithInterceptorChain(RealCall.java:221)

at okhttp3.RealCall$AsyncCall.execute(RealCall.java:172)

at okhttp3.internal.NamedRunnable.run(NamedRunnable.java:32)

at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1167)

gr0uch0mars: navigating to an artist from where?

(I'm not logged in, and searching for an artist)

From home screen. Right after launching it

ok i'll check it

Thanks

zas: Looks like IA is pretty dead right now: https://analytics0.archive.org/stats/s3.php – not sure if there’s anything we can do about CAA queue on our end.

ruaok: hey! Moin.

Freso: yup, issue is clearly on IA side

moooin!

hi pristine__

Was working to utilize bidirectional attribute of AAR. I am not sure if my plan sounds good, wanted your feedback.

shoot

A sec.

First of all, we will get top x artists of the user. Then df1 = get * where artist_name_0 in top x artists.

df2 = get * where artist_name_1 in top x artists

Then for df1 we would call a query to get top y similar artists from each group of top artist s.

And same for df2

Then do union of resulting data frames.

And finally from the union, get top y similar artists from each group of top artists.

We need to combine similar artists from artist_name_1 and artist_name_0 cols and find a collective rank

ruaok: ^

Am I able to explain ?

still waiting on caffeine. hang on.

Yeah sure :)

I know, fetching data from both columns is odd, no?

Yes.

gut feeling... we should work to use one query and union its results into one column.

the overall idea is to use fewer queries in spark.

ruaok, pristine__: what are you looking at?

(what is the bidirectional thing)

in postgres you can do: `SELECT artist_name_0 AS artist_name FROM aar WHERE <....> UNION SELECT artist_name_1 AS artist_name FROM aar WHERE <...>`

reosarevok: the artist-artist-similarities I calculated.

Oh, I see

Yeah, a postgres union seems the simplest then

the values you want can be in artist_name_0 or artist_name_1

pristine__: does that make sense?

Yes, I will have to run and check once though. And one final query to get top y similar artist from each group.

Thanks ruaok :)

np

And yeah, I have submitted the evaluation :)

oh right. time for the fail-all-button. my favorite button ever.

I'm kidding as per usual. :)

Me too. Lol

I've been asking for a fail-all button for 10 years, they never listen to me.

(thankfully)

I've never spent this much time on a GSoC project before. if you think about it, the recommendation project has about 5 people on it.

You, me, iliekcomputers reosarevok and?

alastair?

I wouldn't claim I'm "on" it :p

I just help around because I'm a sucker and bad at not helping

aidanlw17 and you as students.

alastairp, iliekcomputers and myself as mentors/backup.

ruaok: Is that a good thing for you? Having spent so much time. Idk if I should have worked more.

Oh , I forgot aidanlw17 . True

But we will include reosarevok and make it six :)

Because he is bad at not helping :p

50% of evaluation are in!

Only 50% left to fail

evaluations are due by Friday, yeah?

alastairp: yes

:)

28th.

pristine__: I'm happy with how things are going -- the work I am doing has been on my plate for a long time. it feels good to finally knock it out.

<3

ruaok: shhhh, she's not going to believe you're failing her then

reosarevok: I will still come to Spain and sing 'Lalalalalala' in your ears.

That's one (and only) thing I won't to do :p

Want*

mmm, yesterday was monday huh? that kind of crept up on me

gr0uch0mars: fixed it

alastairp: yea, me too.

happy real monday, alastairp.

igualmente

alastairp: to fix the collation mess, I am going to have python do the sorting.

I'll load the data without sort by clauses and then make an indirection array.

drat