#metabrainz

zas: okay, woah, that was faster than I was expecting :) all looks good to me, I assume if something was seriously broken, others would have also pinged me :P

hey shitheads, here's a clue. When you give instructions for the VM and say you should drop the table and import a dump, you should also mention that it totally fucks shit up and how to fix it!!!!!!!!!!

I'm just sayin

yo_momma: the instructions were tested as best as possible by the musicbrainz devs, but it's kinda complex so things can go wrong.

yo_momma: if something does go wrong, coming here and asking politely for help, giving details of the errors that you're seeing, would probably be a better course of action than insulting the devs and complaining without saying what happened.

meh

Nice try, though!

hey guys, you'll all be getting sick of me soon!

Maybe one for ruaok again, but had search server rebuilding indexes

last night and it looks like it's only created one set of indexes, we have recording_index within /home/search/indexdata

http://ec2-52-19-115-70.eu-west-1.compute.amazo... is producing different error now which does make it look like it's working, just missing all the indexes

How did you create the indexes?

java -Xmx512M -jar index-2.0-SNAPSHOT-jar-with-dependencies.jar --indexes-dir /home/search/indexdata --freedb-dump /home/search/<freedb tar.bz2 file>

that command

only change was adding in our db username / password which i changed from the default ones

freedb file was called freedb-complete-20160501.tar.bz2

which at the time of downloading last week, it was the latest one

Try specifying the --indexes artist option

Does it build artist indexes?

nope

well

i don't think so

all it has in

indexdata folder, is "recording_index"

Add that flag, try again.

Not sure why it doesn't build them all.

OK, gotta go.

cool, cheers. enjoy your flight!

No, no months sent. Will try that now.

Uh, obviously not solved. I’ll re-ban, then.

Good day all, I just stood up a RT_SLAVE instance for personal use, and I'm having some difficulty ascertaining how to set up replication when starting the server from source. I see in the vm instructions mention of using bin/replication to run replication tasks, but I do not see this script/executable available in my git folders... is there a wiki or readme somewhere that explains how to get the

bin/replication components built for a source based setup?

paradox606: You need some service such as cron to regularly run admin/cron/slave.sh

chirlu: ok, that makes sense... is there a particular parameter list required or can I just run that script on an interval?

A typical crontab entry would be “8 * * * * cd /home/wherever/musicbrainz-server; eval $( perl -Mlocal::lib ); admin/cron/slave.sh”,

please choose a minute which isn't on the hour :)

A shell started from cron doesn’t execute the .bashrc, so you need to set up local::lib manually.

Yeah, 8 minutes past should be OK.

chirlu: ok, awesome thanks! That is just what I needed!

alastairp, chirlu: ok, I will go ahead and set up for 8 after the hour... you guys rock

thanks!

You’re welcome.

hi hellska

Hi :)

alastairp: I’ll finish the code review of the last PR! Then I’ll start with UUID generation!

gentlecat are u around? just to say hi for now

yep

here again! Alive and kicking!

meh, already having banking issues. I wish vueling didn't waste my time in BCN and I'd gotten all the payments out then.

oh well, life is not meant to be easy, no?

the easy life is where it's at

while vueling did waste my time, they were on time today, alastairp!

bitmap: It’s a bit confusing that beta says it is running the production branch (although the commit ID is that of the head of branch beta).

Not sure what change has caused that.

Oh, the wonders of trying to figure out what I did during my internship and when by looking at github commits

chirlu: not sure how that happened. fixed now

iirc chef sets it something like 'deploy'. so might get reset again...

chef always create and use a 'deploy' branch, which points to whatever you set it in the recipe

at least when using git provider

never really understood why it doesn't just use the branch

I guess so it can support commit ids, tags, etc. in a generic way...

gentlecat: the last commit contains the uuid generation, now the server side part is finished, should I start checking the API endpoints to implement the complete behaviour for submitting dataset items?

mannn

that's really weird to see listenbrainz cited in a paper I'm reviewing

:D

"tl;dr, but it mentions my project, so it must be good"

^^