#musicbrainz-devel

ok, I'm running branch caa on rika now.

is that reasonable?

ruaok: do you have a seperate database on rika? or are the other sandboxes just running with a caa updated database as well?

I'm using the static DB that others are also using.

ok

heh: http://caa.mbsandbox.org/release/7cf16917-2ca6-...

when we do.

I can see where confusion will come from.

nikki: ping

pong

ruaok: yeah. inconsistencies between archive.org and database.

nikki: you following this?

no

ruaok: this is also why we need some form of development environment for this + instructions on how to set it up.

please read the backscroll wrt to final caa testing.

right. seems to be working to submit cover art.

yay, no more port 82. :)

hola

hola

ruaok: why are we using github for metabrainz issues?

I'm using them more for personal notes than official issues, tbh.

ah, ok

nikki: caa.mbsandbox.org is ready for final CAA testing!

(I think)

and yes, the images in s3 don't matter

in ia*

we never do index listings by listing buckets, and we'll never generate an ID that was already in use (unless time somehow starts going backwards)

ah, that explains a few things. good.

then things appear to be working fine.

at least from my cursory glance.

excellent

how does that "Asshole" release have front cover art that's not in the cover art listing?

I'm guessing that is because the caa schema on rika doesn't know about it, but the DB on rika does and it pulls up an older image.

is coverartarchive.org pointing at the right db?

they both point to rika now.

that image comes from coverartarchive.org, not the database though

ocharles hi, could I have your input on chat I had with warp at 7.03pm - 7.38 pm please

ruaok: http://ia600800.s3dns.us.archive.org/mbid-7cf16... the indexer doesn't seem to be running?

maybe because I haven't installed it yet?

but was pgq installed with cover_art_archive triggers?

if so, it'll reindex the moment you launch the daemon

dunno. I followed the instructions.

ijabz: ok

we need a nagios check that pgq and the indexer are running too

a nagios check that the lag isn't huge would be good too

ocharles: you should open an MBH ticket for that with explicit instructions for how djce can test that.

how do I install the caa triggers?

i don't know how nagios works or anything

step 5

ocharles: you dont have to know how nagios works.

ijabz: i don't really have any input right now, would have to look at commit logs to see why that was added

you just need to tell djce how to test for the delay.

he can do the rest.

ok

that's an output of pgqadm

a sample shell script would be of great use.

so hopefully that's easy enough

ok, Ill raise a bug to remind us (though may not be a bug)

i don't think it is a bug

email me instead and i'll look more tomorrow

actually

no harm in a ticket i guess

carry on :)

lol

Yeah, Im not saying its a bug, but better documented in jira

ocharles: the caa indexer readme is a bit conflicting.

Also while your here, couldn't find file or diff for the release group type splitting schema chnages

wait, never mind.

ruaok: also, need input on http://tickets.musicbrainz.org/browse/MBS-4607 as that's entirely by design (when linking to an asin, include the musicbrainz affiliate tag)

ijabz: that might only be on my desktop then

ocharles: ok

we have a report that lists amazon links that need cleaning up, there's really no reason for anyone to care about the amazon url other than to check that the domain and asin are right

nikki: can you please put that into a comment on the ticket?

so as long as the url works, someone like me or salo.rock will clean it up if it needs it

Anyone. is it possible to edit the new track number field in a release yet on test, i tried but nothing happened ?

python hackers: is it possible to remove multiple items from a dictionary with 'del'?

can i del something['foo', 'bar']?

I dont think so.

you could do a list comprehension and throw away the result, that's probably how I'd do it

[del something[entry] for entry in ['foo', 'bar']]

more concise than writing out the loop for real, but otherwise equivalent

I actually only need the del for the side effect

(it's a memcached store)

yay, mutation, yay side effects

del is nothing but a side effect :P

:D

hence throwing away the result

oh, it's not persistent?

(persistent in the sense of functional programming)

not quite sure what you mean

but all del does is decrement a reference count

generally speaking

oh, no, python's not that functional AFAIK

ocharles: del doesn't return anything.

ok

and yeah, that's what I meant by purely side-effect, it doesn't have a return value

you can all critique this python in a moment :)

ocharles: so it would be a bit pointless if it doesn't modify the input "in place"

if it was haskell it'd presumably be :: a -> IO ()

(or so)

ijabz: right now you can only add new tracks, editing the existing track numbers doesn't work

k

meh, i dunno why I don't just write this in haskell, but oh well :P

ocharles: Can you remember what we were talking about recently. About a schema change that was supposed to happen after 6 months?

I've forgotten what it was now :)

adhawkins: disambiguation being misnamed?

i think?

Ah yes, that's right. Got it now

Added the disambiguation element for release group, this is due to replace the badly name comment element but the comment element will not be removed for six months in order to preserve backwards compatability

That was written 9 months ago :)

heh

this is what happens when we don't have tickets :)

Shall I log an issue for it?

please

Against what, server?

yep

Ok

https://gist.github.com/fa44e0d812e237ae41e2 how's my python? :)

ocharles: any reason why you didn't setup logging for caaindexer?

it does have logging

the svc directory is missing the require log subdir.

+d

oh, oops

i think hobbes has that

must have not commited it

pino does.

make me a contributor to caa-indexer and I will commit.

done

thx

Heya!

What do people generally do with copyright notices at the top of source files?

In libmb4 they pretty much all say Copyright (C) 2011 Andrew Hawkins

I can never remember to update the year when I change something :)

I was wondering if anybody can help me install the musicbrainz-server

Maybe someone can

What is the problem?

adhawkins: half the reason i'm quite against them