#metabrainz

do you want me to use the metabrainz OAuth or the authentication token suggested by me

don't worry too much about specifically what option you want to use

I was thinking that we could implement oauth in metabrainz (we have something similar in critiquebrainz)

with the expectation that there might be other things that we want to do in the future on AB that can use it

I'd be happy for your proposal to just say something like "we need authentication, two options are a simple token, or we can implement oauth"

oh okay. So right now I will submit the proposal with the authentication token system and also including the two phase idea for running the jobs. And we can decide on them later.

and if we go ahead with the project, we spend some more time looking into the options and seeing what one is better

and I will mention the two authentication options

yes

Thanks :)

ruaok: sorry for confusion. looks like I was wrong

I was confused about what you were wrong on. :)

everything has to be finished and squared away by Friday evening. no changes. do not pass go. do not collect $200

oh feck. that blows.

I was under the impression that if at least something was submitted by friday, it could be tweaked for a few weeks more

right. see?

why is the student selection phase a freaking month long?

2 months between application deadline and official start

must be all those mega OSS projects ruining it for us little folks.

something to bring up at the summit, I guess

yeah, I can see it being changed after people said the previous one didn't work well

here is the next odd thing.

I can't view a final proposal until the deadline has passed.

wtf?

I guess because we make our community bonding period now, between org announcement and submission

rather than after acceptance

yeah, someone mentioned that on the mentors list. I think they said they might be able to change the app for next year's programme...

armalcolite: I can't see your final proposal. there is one submitted, but I can't see it.

can you please mail a copy of rob@metab.... ?

+it to

LordSputnik: Hey, sorry, I've been slacking with the twitter issues. Thanks for the prod. I'll get on it.

alastairp: I've reviewed the rate limiting stuff on both the server and the scraper and things on the server side look fine.

the client, less so.

do I recall that there are multiple requests happening in parallel on the scraper?

up to 5 or so?

wasn't there some problem with retries?

yes and that is why multiple requests were added, right?

yes, the scraper will send a second request before the first one is finished

up to 10 simultaneous queries

I've yet to find where that code is that fires that off. :(

it's all javascripty

very so.

without looking at the code:

there's some sort of getAndSendNextPage

and somewhere in there is a if numOpenRequests < 10 { do next one }

and numOpenRequests is decremented in the success handler of the submit function

k.

I'll keep digging.

navap: thank you so much, it's really appreciated!

https://acousticbrainz.org/ Total (unique): 1,999,407

smells tantalizingly close to beer time to celebrate 2M. :)

did someone say beer?

\ø

JesseW: just FYI: Discogs and MB are on good terms. At least myself and kevin are. :)

Indeed, Kevin seemed very cool.

Guys, a bit of a random question but is there an easy way to find an artist that was added to the database with a given replication packet?

You could look at the artists updated around that time, but otherwise no.

Well, you could look into the replication packet perhaps, if you have it.

That's what I mean! :)

ruaok: sent a copy.

armalcolite: thanks.

I'm fiddling with the search server indices, so need to find a way to test my changes

is the pdf at official website not accessible?

even i cannot view it.

armalcolite: correct. we can only view it when the submission deadline passes.

but, when i try to open it it opens up a blank tab.

armalcolite: you should've spell checked this. :(

I figured that the easiest way to do it would be to set my server to update a bit after the replication packet has been released, download the replication packet, find a new artist, then search for it locally before and after the replication+index update

ruaok: oh. i sent the wrong one. should i resend?

send me the final one that you submitted.

but looking at the replication packet, finding a new artist might be a bit harder then I originally thought. :)

ok

UmkaDK: why not look a the last_updated timestamps in the aritst table?

keep track of the time you last looked at the artist table.

then examine all rows that have been changed and look for new ones that you haven't seen.

UmkaDK: I’m not sure I understand what you actually need. Couldn’t you just look at the artist(s) with the highest IDs?

chirlu: that would work. even though we keep telling people to not look at the id column. :)

Yep, I think that would definitely work for me!! Thanks guys!!!

Well, not for production use of course.

of course :)

for production use, i would suggest the last_updated timestamps, UmkaDK

kk, thank ruaok

*thanks

Although those don’t distinguish between new artists and artists that newly got edited.

I think I *finally* grokked what is wrong with the rate-limiting branch.

sadly, hugh's improved code is less readable than what we had before.

I think I am going to grab armalcolite's latest code and re-add rate limiting to it.

but now that I understand what is going on, I'll need to have a think on how to add rate-limiting.

I set up a local AcousticBrainz server using vagrant, how can I upload content into the server without dowloading the entire AB DB dump?

you can POST data to the api endpoint /<MBID>/low-level

hellska: you can use the acousticBrainz client

and just edit the config file to upload it on localhost rather than acousticbrainz.org

I tryed to post the data using the API but i get an error

what's the error?

400 BAD FORMAT

do you have the acousticbrainz client?

yes I downloaded both the gui and the command line

I posted to the endpoint using python requests

well i dont know about the bad format error

its better to simply use the acoustic brainz submission client

just modify the file default.conf

ok I’ll try to set up it now!!! thanx

when you are done setting it up, just modify your default.conf

have a look at the gist

Yo!

hellska: Tu qué?

jajaja YO!!! is hiphop slang to say ‘cool’ or similar ;)

Sí, yo sé. :)

hellska: were you able to upload the files?

running the client in this moment

oh cool

kartikgupta0909: Yeah uploaded!!!

nice :)

I can see from the server log, but there’s a way to refresh the website view to list the files!

it gets refreshed i think in every 10 minutes or so

or have a look at the database

run psql - U acousticbrainz acousticbrainz