#metabrainz

Oh, Monday.

ruaok: Ping? Home yet?

Project acousticbrainz-server build #78: SUCCESS in 5 min 10 sec: https://ci.metabrainz.org/job/acousticbrainz-se...

Freso: pong. No.

ruaok: Oh. Mailed you instead.

Basically just a heads up about #metabrainz meetings the coming two weeks.

Hey ruaok: should the search server be working ok using commit 4f677727 from mmd-schema and latest from search-server? I can't seem to load the indexes with jetty or tomcat

I think this is the error: "com.jthink.brainz.mmd" doesnt contain ObjectFactory.class or jaxb.index

iggleby: yeah, there seems to be some issue. the old war should still work fine: http://ftp.musicbrainz.org/pub/musicbrainz/sear...

gentlecat: the client for dataset submission is completed(https://github.com/MTG/acousticbrainz..., I think it’s time to test/work on the Dataset API, can we speak about this before I start?

thanks ruaok

Moin

moin, zas.

in order to try and pry my brain out of hippiedom, I'm going to try watch the docker videos.

hello hippies

Hi alastairp! Still not able to change the binary in the gaia2 library to be uploaded to the PPA. Unfortunately their patch mechanisms do not support binary changes.

ah, that's a really reasonable decision

no problem. I'll speak with the gaia maintainer today to get your patch merged

oh cool

how about we put gaia to the side and start working on the next part

yes

writing the server code

could you make a branch in the repo for this project

you have push access to the repository, you should be able to do it

oh cool

I ll start working about giving an option to the user for running the task on his own machine.

why have a branch and not merge with master?

Gentlecat: Until its complete most of the features wont make sense

as chirlu was saying before, there's a history of never releasing things

like what will the user do of the option of running the job on its own machine when you cant do that for the moment

Gentlecat: ? but a PR will be on a branch until we merge it

feature flag?

perhaps we're all thinking about a different way of doing things

there's a lot of work on the server going on in parallel

it might be a nightmare to merge all of that

kartikgupta0909: let's work out the next things that we're doing and open some tickets for it

sure

but until it becomes a nightmare, let's not make up new ways of working

according to the proposal it is giving an option to the user and authentication token

not to mention that there are still a lot of things that aren't moving anywhere https://github.com/metabrainz/acousticbrainz-se...

kartikgupta0909: so, as before, get the next thing on your plan, and break it up into small bits of work, of no more than 1 week

sure.

1.Giving an option

kartikgupta0909: Gentlecat has authentication working here: https://github.com/metabrainz/acousticbrainz-se...

so do we wait for this to get merged for the authentication part?

I'm reviewing patches now

One thing to alleviate the pain of merging a feature branch into a constantly evolving master branch, is to frequently merge the master branch into the feature branch (or rebase the feat. branch on top of latest master... I prefer the merge though, but YMMV).

oh cool. Until then I can work on the giving option part. If the authentication is merged until I am done with that then we will use that otherwise I ll wait and maybe do something else. Also we will have to do schema changes.

kartikgupta0909: yes, good idea. Please consider sketching out how you think this interface might work (you did this a bit on your proposal)

yes sure.

it would be great to see the ticket include this information: 1) a brief overview of what the ticket will do, 2) a description, or sketches, of how the workflow will work, 3) the beginnings of a list or what things you think you need to add or change

especially in the part 3, I want you to specify if you need to add new tables to the database, or change existing tables

On it .

hellska: I didn't finish implementation of the API

and don't know if I will in time, have a lot of other things going on atm

I can help on that if you want!

Seems both the main mirrors have not upgraded the schema yet so no replication

hellska: sure, see how much time this will take

so that you don't get too far off track with the project

gentlecat: At least on the endpoint to submit recordings without MBID that is connected to the GSoC project

I'm not sure if it should be a separate endpoint

ok … but the current endpoint needs an UUID to submit lowlevel data

oh, you are talking about data submission

yes!

I don't know what's the point in having it there since it can be taken from metadata

and in your case if it isn't, then we can generate a temporary UUID

So you mean that the uuid is passed as an argument instead of a part of the URL? If it’s present fine if not the server tries to find or generates a new one!

what uuid? mbid?

yes the mbid!

it's passed with the data so we don't need to duplicate it in a request

this looks unnecessary to me

…oh ok got it, the MBID/uuid is read from the metadata, if it’s there the server will use it if not it search or generates a new one.

I will figure out how to implement those changes, and reflect them into the client as long as I finish!

maybe it's better to focus on the server first

you can implement all the features in the client, but they aren't going to work unless server has support for them

and it can be hard to predict how the interface will look like before it's done

Yes, is exactly what I mean, work on the server side first!

so it's probably better to implement the API and then add support for these features in the client

Yo!!! I’ll go for it, I let you know if I get stuck on something

:) thanx

alastairp: http://tickets.musicbrainz.org/browse/AB-132 , have a look when you get time.

Project acousticbrainz-server build #79: SUCCESS in 4 min 57 sec: https://ci.metabrainz.org/job/acousticbrainz-se...

thanks