#metabrainz

[listenbrainz-android] 14prabalsingh24 opened pull request #161 (03main…dont-fetch-url-twice): Fix: don't fetch cover art urls twice https://github.com/metabrainz/listenbrainz-andr...

mayhem: (replying to something from way back) - that is clearly a keyboard

jasje: akshaaatt: icon changes, exiting! Hopefully sticking to the ones we discussed :P

Also, remember that in the style guide we avoid using 100% black, we have a different ‘black’ colour

reosarevok: when you’re online can you check the report I sent? Would be good to have it looked at soon, and have the community managers input before re-allowing editing rights (let me know if you can’t)

Hi prabalsingh! Thanks a lot for your contributions :) We would love any help we can get for our projects. Feel free to pick up bugs and submit a PR for them. Let me know if you need some help understanding anything or fixing any issues.

aerozol yus for sure

[listenbrainz-android] 14akshaaatt merged pull request #159 (03main…Fix-theme-switch): Fixed theme switch https://github.com/metabrainz/listenbrainz-andr...

Thanks for bringing up the listens issue prabalsingh! It is weird as to why that api is taking so long for you.

lucifer could you please look into the issue here? https://api.listenbrainz.org/1/user/prabalsingh... seems to be taking forever to load. Is there a reason behind it?

akshaaatt: we should probably use viewpager for this'

Yeah. Infinite loading will be the key

You are already making it for feed, right jasje?

yes

Then you would be the right person to update it for profile as well

but i need to edit the endpoints themselves first

akshaaatt: or

Not the endpoints but the params which go in

iss complicated

zas: is 15 UTC ok for the database downtime?

It's not. I have the code for things if you want me to work on it.

akshaaatt: but anyways i see you removed preloader from listens screen

Yes that was causing a crash

akshaaatt: im talking about feed

There seems be to be some fault from glide

akshaaatt: its kinda restrictive but works good when it does

coil is so shit

akshaaatt: since we have removed the preloader

akshaaatt: load the listens in smaller batches. `https://api.listenbrainz.org/1/user/prabalsingh24/listens` will give 25 listens at max by default. look at the listened_at value of the last listen in that list and then next time query with that as `max_ts`, `https://api.listenbrainz.org/1/user/prabalsingh24/listens?max_ts=1685294816` so on.

might as well add the fallback cover art innit?

zas: anyway I prepared a google doc & invited you & others

O’Moin

bitmap: ok for me

monkey: up?

lucifer: ?

just added the test to the new endpoint

also wanted to know

what max limit do you suggest for getting follow listens?

current is 2

per user

and 10 overall

Cool lucifer. We'll do that soon. Just wanted to confirm that the api had no issues

moooin!

aerozol: key-board. ha, love it!

Moin!

after this weekend its going to take a bit to readjust to work -- spring in BCN is really lovely with loads of stuff going on.

Life is hard when life is good!

yvanzo: https://github.com/metabrainz/musicbrainz-serve... seems ready to merge I think

Do we have anything else to put on a beta yet?

[musicbrainz-server] 14yvanzo merged pull request #2953 (03master…cldr-43-mbs-13102): MBS-13102: Update alias locales to Unicode CLDR 43 https://github.com/metabrainz/musicbrainz-serve...

No, let's release ;)

I'm ok with that, let's put this into beta for now

Remind me

What happens in prod when someone uses one of these scripts in beta?

*locales

Do they just show as missing a locale but otherwise work fine?

I’m joking.

Good question, I’ll test locally.

I'm not, I think it'd be good to put this into beta so CatQuest can start fixing aliases :p

👍

Assuming it won't break anything in prod, anyway :)

I don’t think it will ISE at least.

[listenbrainz-android] 1407jasjeet opened pull request #162 (03main…feed-section-phase-1): MOBILE-137: Feed Section Phase - 1 https://github.com/metabrainz/listenbrainz-andr...

I'm reviewing your core dorps

why can't i resolve commits of my own branch wtf

*drops

conflicts*

"yvanzo/refactor-prc" are you fixing China for us too, yvanzo? :D

It stands for PurRing Cat.

jasje: Are you thinking about limiting the amount of songs each followed user can contribute to the follow feed?

The label thing I forgot to drop with https://tickets.metabrainz.org/browse/MBS-12279 - I'll add a commit to your PR amending it

MBS-12279: Drop the manual label removal option

aerozol: its already limited

im going to raise the limit

Hold on, for "processing replication changes" actually.

jasje: It’s limited in the other feed, sure

yvanzo: oh, ok :)

If it’s limited we might need to reconsider the UI/UX jasje, to indicate that people aren’t getting all the listens of their follows in chronological order (which I would assume)

aerozol: current feed shows only 2 listens per user and total 10 listen events only

jasje: Yup, no point in making the same feed in another tab though, right!

yup i was thing 4-5 per user and 25 as the limit

was thinking**

the frontend can then fetch next 25 by using min_ts of present list

yvanzo: if you could check small bug fixes https://github.com/metabrainz/musicbrainz-serve... and https://github.com/metabrainz/musicbrainz-serve... it would be great to merge those too, and then put a basic beta out

jasje: That sounds ok. The less of a limit we can have the more useful it is though, imo

But if you don't have the time I'm ok with putting this out first and those later on :)

aerozol: yeahh

Interesting how this is apparently hard on server load, but we can do global stats? I guess we don’t update those that much

aerozol: we are same for the most part

xD

phew conflicts resolved

Wow! MB made the top 7 😅

"buy"

What?

Yeah. Also, Discogs? Same

I expect chatgpt wrote this :p

Just some SEO Chat GPT-written page. I won’t give them the satisfaction of a link

Can’t believe Discogs beat us!

[musicbrainz-server] 14reosarevok merged pull request #2946 (03master…MBS-13088): MBS-13088: Standardize series type bubble https://github.com/metabrainz/musicbrainz-serve...

zas: I tried MOVEREPLICA but got replica "does not exist"

Also tried setting maxShardsPerNode to 1 (currently -1) for annotation but it didn't change.

Same for autoAddReplicas (is false instead of true).

My bad, it's about the number of shards, not the number of replica.

nrtReplicas is set to 1

[musicbrainz-server] 14reosarevok merged pull request #2949 (03master…MBS-12211): MBS-12211: Avoid crash when URL has been removed https://github.com/metabrainz/musicbrainz-serve...

Thanks yvanzo

thanzo

jasje: hi! I'm online now

autoAddReplicas is dropped from Solr 9 anyway, not going to use it.

monkey: about the listens section in feed

Just reading the backlog now. I agree with aerozol, I think for this feature of your friends' listens, maybe we shouldn't limit the number of listens, and instead just show listens chronologically

We limited it originally in the feed so that it wouldn't take over other feed events, but I'm not sure that's what we want if we separate the feature

alright

if you still wanted to limit somehow

so that a single user cannot just take up the whole listens section because their listens have the timestamp priority

Sorry, i'm not sure of what you're asking

I understand we could have cases where one user's listens takes most or all of the page, but I think that's more logical (to me at least) than arbitrary limits. As aerozol said the UI would need to be different in ways I don't really know yet if we wanted to do that

It makes pagination more complicated, and probably not reproducible

as in moving form one page to another and then back, you might not have the same page

zas: DELETEREPLICA gave a more useful error message: the replica identifier is "core_nodeX" rather than core name "annotation_shard1_replicaY"

So moving annotation replica worked now :)

Added a url replica through Web UI.

[musicbrainz-server] 14yvanzo merged pull request #2954 (03master…nail-last-core-types): MBS-12552 (pre): Nail last "core" Flow.js types https://github.com/metabrainz/musicbrainz-serve...

monkey: so we are settling on no limit