#metabrainz

resarevok, yvanzo, zas: I put the event-art-archive branch back on test.mb, which is what was running when we had those OOM events. if they happen again I'll do a deeper investigation

reosarevok*

bitmap: it just happened again

yep, that was fast

akshaaatt (IRC): I had submitted my proposal on the GSoC site 3-4 days ago.. after u told to add the project size in my proposal, I tried editing the proposal I uploaded on the site but it’s not working for some reason.. I hope that’s fine as there was an option to select project size on the site along with the application..

reosarevok, yvanzo, zas: the OOM issue should be fixed now (inb4 telegram alert)

That’s fine pranav[m] . No worries!

Also akshaaatt (IRC) some minor things are still left in overview screens shud I add them in this PR itself? It may go upto 2k ig.. I’m not sure

Go ahead and add them in the current pr. I haven’t gotten the time yet to review it

pranav[m]: what issue are you facing on the GSoC website?

I had uploaded my proposal on the site .. when I am clicking on edit to update my proposal, it takes forever to upload the new proposal.. it just keeps showing a circular loading progress indicator widget .. lucifer (IRC)

pranav[m]: try logging out and logging in, clearing site data/cookies and trying again.

Alright I’ll try that once thanks a lot lucifer (IRC)

I tried that but no luck lucifer (IRC) .. I’ll try changing devices once.. I think the problem is with googles uploading service tho.. so I don’t think changing devices will help

pranav[m]: I see, send an email to the GSoC admins with the issue. probably include a screenshot as well.

Yup, will do

mayhem: okay will do in the next few days. We have a long weekend here. Scope is just that use case or my app as a whole or all my thought on playlist generation at all?

With "playable everywhere" you mean what exactly?

Eryn_FL: I have a bunch of playlists, one for every artist i like, containing each song only once. Then some mixed playlists. I have a little tool which copies playlists and its content to any place. When my library, some playlists or the metadata changes, i use that tool to just copy/overwite everything on my target destination (phone). Is the last step what you want?

lucifer (IRC): thanks it got resolved!

[20:13] <bitmap> IIIRC it only left like 6k editors

seems reight to me, sleeper spammer account number in the several thousands when we became aware of it some, idk 10ish years ago

pretty much

lucifer: workin on LB-1459, I have setup the developemnt environment, added websockets to test container, verified it is working with docker ps command, added python-socketio to dev requirements. But now I am getting errors when connecting to localhost:8102 in test file. I tried independently connecting to the same in jupyter notebook, it worked perfectly, any idea what could have been the error, I am stuck at this point

LB-1459: Add integration tests for websockets backend https://tickets.metabrainz.org/browse/LB-1459

munishk: what's the error?

<07munishk> Cannot connect to server

<07munishk> I am using self.sio.connect('http://localhost:8102') The error is: socketio.exceptions.ConnectionError: HTTPConnectionPool(host='localhost', port=8102): Max retries exceeded with url: /socket.io/?transport=polling&EIO=4&t=1711608793.9651434 (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7fd9ec337150>: Failed to establish a new connection: [Errno 111] Connection refused'))

munisk: if you can share the code on GitHub, I can try to debug it locally

Okk

!m bitmap

You're doing good work, bitmap!

Done with the community post. ＼（＾ ＾）／

@yvanzo thanks for saving my ass twice!

I have tried to make smaller goals this time. I hope this is fine now, but if you feel anything is missing, please do reply on the post!

lucifer: I have uploaded the code on github: https://github.com/munish0838/listenbrainz-serv...

mayhem: I have tried updating the proposal. I hope I was able to make the relevant changes: https://community.metabrainz.org/t/gsoc-2024-bp...

let me look in a bit.

relaxoMob: scope is pie in the sky for the festival app -- not making any promises that we can implement it all, but it would be good to know what you think is best. for other playlist suggestions, please add a different ticket.

hehehehheh. :)

mayhem: ok, will dream a little. You mean ticket in Jira right? Not the forum.

correct.

The meme depends on if we see the parents as developer or as another stakeholder

atj, bitmap, reosarevok ^

yellowhatpro: no exaggeration please, you took a lot of initiative and have been very responsive to my and atj’s feedback.

munishk: try connecting to the websockets container at `http://websockets:8102`

[design-system] 14dependabot[bot] opened pull request #125 (03master…dependabot/npm_and_yarn/express-4.19.2): Bump express from 4.18.1 to 4.19.2 https://github.com/metabrainz/design-system/pul...

<07munishk> lucifer: on it! will update

yvanzo: Hey! have you read the fluent docs? I was just going through them

So the reason why you prefer it more is because of the attribute feature it has I presume. More details here : https://projectfluent.org/fluent/guide/terms.ht...

It can be used to describe specific things in grammar such as gender and plurality which are available in i18next json as well. What was the case you were saying it considers that is required for relationships?

https://musicbrainz.org/relationship/a3005666-a... - yvanzo you provided this as an example but I was unable to get what you were trying to show exactly

hi aabbi15: see the link phrases in this page

yvanzo: do you know if there any example MB Solr update payloads I can use for testing?

I think I have a working Solr 9 cluster with mb-solr and schema loaded

need to test submitting some documents

aabbi15: it is converted into this entry in Gettext: https://github.com/metabrainz/musicbrainz-serve...

ah, looks like the sir test cases might be usable

aabbi15: It lands in Weblate as this source string: https://translations.metabrainz.org/translate/m...

aabbi15: It would be much easier to express it in Fluent format, and it would allow to generate better phrases

aabbi15: better as in grammatical correct at least, and even more natural

atj: using sir yes, usually I set up a sample database and I rebuild a small index to start with.

aabbi15: Here is another example that is not even grammatically correct in English: https://musicbrainz.org/relationship/5be4c609-9...

aabbi15: The “n” will be missing in front of “original” or “eponymous” if it isn’t an “additional” member.

aabbi15: Also there is the tense “is/was” which could be refined depending on the dates of the relationship.

(or just depending on the “ended” boolean flag.)

Hey lucifer (IRC) and monkey , what's you guys' suggestion on the project size for the Subsonic Integration project?

yvanzo: 🥳 https://www.irccloud.com/pastebin/jBFV5BsE/

2 node Solr cluster with configuration stored in a 3 node ZooKeeper cluster all deployed using Ansible

we're not far off being able to deploy a test environment

<07munishk> lucifer: I get this for websockets:8102 -> Temporary failure in name resolution socketio.exceptions.ConnectionError: HTTPConnectionPool(host='websockets', port=8102): Max retries exceeded with url: /socket.io/?transport=polling&EIO=4&t=1711633216.7422407 (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7f4e9df2e4d0>: Failed to establish a new connection: [Errno -3] Temporary failure in name

resolution'))

!m atj

You're doing good work, atj!

munishk: I find the timeline very confusing. I think you should mark each week clearly with a date of week of month. right now its hard to follow.

and the section for community bonding period 1 may - 26 may has tasks that should take all of 2 hours work -- there are very few things to decide.

"sequntial" and probably spell check your document. :)

<07.zima_blue> Regarding the memory requirements for a mirror server, I see "RAM: 16 GB (or 4 without indexed search)." Is 16 only needed if you're building the search indexes manually or do you need that much to use the indexes in general?

remove: "Setting up server to save the generated bpm values (and confidence scores)" from week 3/4 . I'll do that instead.

"Completion of Phase 1 and starting Phase 2" useless noise. remove.

"Obtaining the bpm values for songs from Spotify" don't bother. I'll do that on the server.

Hi zima_blue: it is recommended in general but it can run with less.

<07.zima_blue> Hi yvanzo: Thanks for the info! I was doing a cost comparison between 16 threads with 16gb ram & 2 threads with 4gb ram and the difference is pretty substantial ($400 vs $40). Does that sound about right?

<07.zima_blue> *AWS EC2 cost comparison

friends dont let friend use AWS, _zime_blue. Use hetzner instead.

zima_blue: it depends on your usage for this mirror, but yes it seems excessive.

<07.zima_blue> Thank you both. I'll do some more research. Do have an idea of what kind of costs I should be expecting with both of those options I mentioned?

zima_blue: a Hetzner shared CPU VPS should be sufficient, maybe a CPX31 (€13.60 pm) or a CX41 (€17.40 pm)

actually CPX 41 because of the disk space requirements

(€25.20 pm)

yeah, that will be probably be cheaper than adding more storage

didn’t know it was possible 😅 you know better!

i wasn't thinking about storage to be honest!

i forgot that the search cores needed a lot of space

(and it's just the beginning!)

i started looking at the custom tokenizers we use and i get the feeling they should probably be reviewed / updated

not sure we'll have enough time though

<07.zima_blue> Thanks everyone! Much appreciated : )

got it with fluent we can add any sort of custom variables as attribute

SO using it would be better

yvanzo: Thanks a lot! You have been very helpful regarding this. I will integrate fluent in my proposal as well

reosarevok: yvanzo: I wrote a rough outline for deploying the artwork-indexer container so that I can proceed with the EAA - https://gist.github.com/mwiencek/1fead215ac810e...

if that makes sense I can work on it today or tomorrow

aabbi15: do you have an example using attributes?