#metabrainz

weird, and a pity about not spain though :(

but ticektmaster using mb data is hella great :D

Leo_Verto: also, you mean Finns. "fins" are the ones that move a fish through the water. "Finns" are the ones that move vodka through themselves

lmao

mayhem: ping, I was hoping you could tell me what work you've done in your big-query branch on listenbrainz and what remains, so maybe I could help out with the branch a little.

sure.

kinda bored of fixing minor bugs :P

the main point of work was to shovel the data to bigquery.

but, bigquery cannot dedup the incoming listens.

and postgres isn't a long term viable data storage.

Heh, I don't think I want to go there, but nice poster. http://www.nummirock.fi/

based on positive feedback on influxfb, I decided to use influxdb as a way to keep the master archival copy of the data and de-dep it at the same time.

and kafka sucks ass. anything java related ends up being a pain in the ass.

so I tossed that out as well.

so the bigquery branch has a misleading name, overall. it should be "big-ass-overhaul" really

I replaced kafka with a custom redis construction.

in mspaint?

mayhem: if I understand correctly: user -> listenbrainz -> influx for dedup -> bigquery is the plan? please correct me if i'm wrong

pretty much.

the influx bubble is the master archive.

mayhem: okay, makes sense.

great.

now, what is left?

the test system pre-dates the docker files.

so the test system needs to be moved to docker, which is where I am now.

are you docker versed?

mayhem: don't have that much experience but if you point me to what to do, I'm sure I could do it.

ok, that is one option. but lets explore some more.

the other thing that needs doing is import from listenbrainz (alpha).

there are some people who have been scrobbling to LB, even in its alpha state.

yeah i've been scrobbling too

the current DB is inaccurate due to the second resolution bug from the last.fm web pages vs the last.fm API

so, migrating the DB is not a good idea.

instead what we need to do is offer an option for someone to import from the LB alpha version into LB beta version.

which entails reading all of the listens from a given user and importing them into LB.

see what I mean?

mayhem: reading from where? the current listenbrainz db?

correct.

okay, that seems doable

at some point when we release this, I'll move listebrainz.org to alpha.listenbrainz.org and release the beta version under listenbrainz.org

so, if you would like to contribute new code, writing this import function would be a really great asset.

mayhem: just so I'm completely clear, what we need is an option in the beta that reads from the current postgres db and then sends it to influx, yes?

I would set it up on a higher level.

meaning that I would do it on an HTTP basis.

it could even be a separate web app, really.

fetch all the listens from one and submit to the new one.

a seperate web app that fetches from the listenbrainz api and then submits to the beta?

yes.

okay

or we could include it into the LB app -- doesn't really matter too much.

mayhem: but how do I submit listens to the beta right now?

I guess including it is slightly easier for users to understand.

you can't, since that service isn't up.

but, you can start your own copy.

checkout the bigquery branch and run it.

right, indeed.

mayhem: I think I get it, I think I'll start with a new app and then we'll see about moving it into listenbrainz

time to up the ante a bit.

as long as they are both flask apps, that should work.

and as I said earlier, I am out this week.

but hopefully next week I can put my head down for a few days and get the BQ branch ready for review.

enjoy your holidays =)

mayhem: that's okay, I'll probably need a week to get it done anyways

Have fun!

iliekcomputers: perfect.

mayhem: thanks for the help.

holidays. :) I'm going to a 1.5 day conference in Chicago.

what a stooopdily long way to go for 1.5 days. :(

what conference is it?

> 250 hand-picked attendees

seems exclusive :)

buut chicago!

hard to get into, yes.

but worth the trip. even in jan-fucking-uary!

Musicbrainz: would it be possible to make a checkbox for "video" on recordings when you edit a release?

mayhem: just a small question, asking for the listenbrainz user token in external apps isn't a security no-no, right? the token is made to be shared with apps?

totally fine.

and dealing with this is probably a good reason to not have a separate app. just keep it in one... makes it easier for the user too...

SothoTalKer: IIRC there's already a ticket

yeah, you're right.

just base your work off the bigquery branch.

okay

anyone knows what https://github.com/mineo/mb-rngpy is?

musicbrainz relax ng python something your mom?

it's a random number generator, of course :D

looks like there's one more spare

ohh, you dig yer mom jokes?

I need someone to enjoy those with.

of course!

who doesn't?

loads of people here only know them because of me.

but they don't really *know* them.

except reosarevok's mom. she seems to get it.

LOL

check your sent mail folder and see if a convo matches?

and yes, this is not an exact science where you turn over product after payment.

come on.....your mom jokes are a solid portion of important history that should be referenced as often as possible....

Quesito: <3

I <3 and your mom.

I <3 your mom too.

I got nothing--nothing is adding up with this cat.

I'm just digging as I don't buy what he's selling.

but as he used my name....it's recent?

ask him what his product is

must be. I was kinda surprised to see it him mention you.

yeah--no matching names, websites, email addresses....nothing.

email him and start a convo. see if we're being paranoid or if he remains cagey.

I put a message on the communtiy thread--but will email as well.

someone should give Quesito "MetaBrainz something something" titles on community.meb

+a *title

arrrr

Mineo (and Quesito): done!

I have a title now? I just got my first badge...things are moving so quickly! :)

I see it's one of the dependencies for SIR

*ugh, got disconnected*

Mineo: https://github.com/mineo/mb-rngpy ^

can you clarify what this does?

it's basically the mmd-schema package for python

it's a bunch of classes for all the things defined in the xml schema that can be serialized to xml

ah

so it needs an update after the last schema change, right?

and can we move to metabrainz org or is that something you built on your own?

we can move it

Gentlecat: can you create repositories in the meb organization? I need to transfer it to someone who can do that first

yes