#metabrainz

[listenbrainz-android] 14dependabot[bot] opened pull request #189 (03main…dependabot/gradle/org.jetbrains.kotlinx-kotlinx-coroutines-test-1.7.2): Bump org.jetbrains.kotlinx:kotlinx-coroutines-test from 1.7.1 to 1.7.2 https://github.com/metabrainz/listenbrainz-andr...

Just popping in to drop this, had an alert for some keyqords: https://hydrogenaud.io/index.php/topic,124384.m...

*keywords

Foobar2000 ListenBrainz plugin, looks pretty cool. And has recommendations for artist/track etc?

mayhem: FYI I am honoured that MeB will consider flying me across the globe, and if finances don’t allow for it I will not take it personally :)

ricksucks: and lucifer: I went through the doc Lucifer shared the other day and it gave me some idea but it would be great if you guys could give some pointers on how exactly can I implement it.

riksucks ^

riksucks:

aerozol: I did finances, we can afford it. we should look for tickets for you soon.

[listenbrainz-server] 14amCap1712 merged pull request #2513 (03master…bulk-spotify-export): Export batch generated weekly playlists to Spotify https://github.com/metabrainz/listenbrainz-serv...

lucifer: can you please put the tag query stuff back on test?

sure

thx

mayhem: do you want to release the weekly recommendations stuff?

yes, please. (but not monkey's branch just yet)

once the generation of the playlists works as we expect, we merge monkey's branch and set the crontab to weekly.

there's also one issue about that. currenlty those playlists don't take timezones into account.

i think the solution would be to run the cronjon hourly like we do for daily jams and pass an explicit list of users to spark to generate playlists for

oh. hmm.

i'll add that

yes, please.

but lets release the just merged PR to make sure that works...

i tested it already, it works.

mayhem: done for tags

you need to provide a condition always for now. even if its one tag.

pass multiple tags by adding `tag=:tag_name` each time

ok, thanks. is the : correct there?

or is that the placeholder name?

its not -- ok, cool. :) thanks let me send invoices and then I'll dive it.

in

[listenbrainz-server] 14KasukabeDefenceForce opened pull request #2514 (03master…build-problem): Updated dependencies. https://github.com/metabrainz/listenbrainz-serv...

mayhem: yeah :tag_name is the placeholder

mayhem: <3 can’t wait!!!

yvanzo: seems decoda.com (which we banned as malware) is now the site for some mining-related company

Should we unmalware those I guess?

(starzik is still dodgy)

So it’s time to invest in cryptos?

The physical sort :D

But yes, it’s completely fine to drop it now, it’s been banned for 2 years, I didn’t even remember it used to be a lyrics website.

It’s not listed in URL reputation databases either.

mayhem: https://usercontent.irccloud-cdn.com/file/1IiaG...

*chef's kiss*

Now to see if it actually works :p

lucifer: anymore changes in feed following?

yvanzo: ok, pushing the drop commit to master - seems pointless to have a PR for it. Will explain the issue in the commit message :)

Thanks for checking!

jasje: will check and let you know

99% of which will be unread by the people who are supposed to read it.

for sure.

lucifer: really hope you get in this time 🤞

reosarevok: thanks! hope so

[listenbrainz-server] 14amCap1712 merged pull request #2483 (03master…separate-follow-listes-lb-1269): LB-1269: Separate follow listens from feed https://github.com/metabrainz/listenbrainz-serv...

jasje: thanks for the contribution, merged.

vscode_: hi! ping me when around and i'll walk you through on the changes required.

arsh: ping

great onto similar listens now!

Hi mayhem:

hi!

do you have a day job or something arsh? you should be more available to be working on your gsoc project....

No

I believe its the time difference

I do work sometimes on weekends though

Sure I will allocate more time to it

I've been available in the evenings, so that shouldn't be it. I work with others on pacific timezone and its not a problem.

man the Schengen visa thing is bullshit

if i wanted to go to india i bet there wouldn't be any issue

yes, you're quite behind and I would really need you to show me in the next few days that you can catch up.

once behind, most students never get caught up.

Yes I agree I have been taking a lit longer than I expected as well

Sure I will try my best

that said, before the mid-term evaluation you'll need to open a PR against the LB codebase.

ok

for now, it is fine to check-in the codepen (or whatever) code you've written. there is no need to integrate it wit the LB site just yet.

but, I need to see your code and see a PR, before I can pass you. ideally, have the PR be approved too.

Sure

I will do it at the earliest

ok, great.

mayhem: After mergin master in my feature branch, running the web docker container exits early with this error: `ModuleNotFoundError: No module named 'troi.patches.periodic_jams'`. Tried rebuilding the image, no luck. Anything I should be doing?

Full tracelog:

monkey: tried `./develop.sh build` ?

Yes

Same result

let me check

then what does requirements.txt say about troi?

`troi@git+https://github.com/metabrainz/troi-recommendation-playground.git@ad2cf2fe2486efc5f9c40b5456ec7be0d949cf42`

try @main

and then build again.

OK, will try tomorrow

[troi-recommendation-playground] 14amCap1712 opened pull request #104 (03main…spotify-export-util): Separate spotify submission into a utility function https://github.com/metabrainz/troi-recommendati...

mayhem: ^ mostly moving code around

mañana, mañana, nur nicht heute, sagen alle faulen leute.

lucifer: yeah, makes sense. LGTM

[troi-recommendation-playground] 14amCap1712 merged pull request #104 (03main…spotify-export-util): Separate spotify submission into a utility function https://github.com/metabrainz/troi-recommendati...

One question mayhem: Are there any particular steps for the submitting the PR and any specific directory where I should submit code

this might help, monkey had shared a step by step list on how to add a new page to LB: https://chatlogs.metabrainz.org/libera/metabrai...

not any particular steps, per se.

lucifer: we're not quite ready to add that page yet. we're just adding the contents to a codepen for the time being.

mayhem: just copy the contents from the codepen to a js/html component like that?

maybe don't add the flask route.

I was thinking of temporarily dumping everything into frontend/js/src/explore/artist-similarity

and don't even connect anything to the rest of LB just yet.

more work needs to continue in the codepen, but for review purposes we want to have it in the LB repo.

arsh: does that make sense?

Yes understood

cool makes sense

mayhem: I believe I should leave the tinkering of the graph and start working on layout of the page

[troi-recommendation-playground] release 03v-2023-07-05 has been published by 14amCap1712: https://github.com/metabrainz/troi-recommendati...

Does it sound good to you?

After submitting the PR

arsh: yes, exactly that. Time to move on to other bits.

ok thanks

lucifer: I am here, can you please walk me through steps as you mentioned?

vscode_: hi! so for the music services page mentioned in step 1 of that document, you need to add a section here: https://github.com/metabrainz/listenbrainz-serv...

ok

2 buttons: connect and disable. when clicking on connect, trigger a request to LB webserver like: https://github.com/metabrainz/listenbrainz-serv...

I will open a PR that automatically generates a new developer token, the endpoint you add should call that.

Look at the https://github.com/metabrainz/listenbrainz-serv... to see how we store tokens in the database

that automatically generated token has to be user specific and should be stored in the table.

finally inject the token in global props here: https://github.com/metabrainz/listenbrainz-serv...

you can see how we insert tokens for other services as well.

let's get this part rolling first and then work on the steps on how to integrate music user token

does that make sense?

ok got it

thank you!

Hi jasje

I have added a review for your search PR

akshaaatt: yo!

i see

And tbh I'm unsure as to why it went too lengthy

We should work on this

?

There's a lot unpacking in the search PR