IRC log of musicbrainz-devel on 2012-06-25

Timestamps are in UTC.

00:41:14 [kurtjx]
kurtjx has joined #musicbrainz-devel
00:52:56 [ocharles]
ocharles has joined #musicbrainz-devel
00:57:42 [kurtjx]
kurtjx has joined #musicbrainz-devel
03:52:23 [alastairp]
hmm
03:52:52 [alastairp]
ocharles: warp: does the test database have a fixture, with sample releases, etc
03:53:31 [alastairp]
I need releases to add to a collection, but would prefer to not need to add releases in setup too
04:20:17 [ianmcorvidae]
ocharles, warp: additional pestering, how do I even set up a test DB?
04:20:27 [ianmcorvidae]
well, alastairp, you may know I guess, since you're doing testing :)
04:46:57 [alastairp]
ianmcorvidae: yeah
04:47:08 [alastairp]
you need a test stub in your dbdefs
04:47:42 [alastairp]
to start with, change the database name in READWRITE to musicbrainz_test
04:47:45 [alastairp]
then run initdb
04:48:01 [alastairp]
then revert READWRITE, and add a TEST that points to that db
04:49:48 [ianmcorvidae]
cool
04:49:54 [ianmcorvidae]
and then run tests, or are there more steps in there?
04:50:39 [ianmcorvidae]
ah, looks to be working
04:50:39 [ianmcorvidae]
cool
04:50:45 [ianmcorvidae]
er, maybe
04:51:15 [alastairp]
HACKING says how to run tests
04:51:29 [alastairp]
I can run some of them. Some fail for me. not sure why
04:51:36 [alastairp]
but anyway, it's late
04:52:09 [ianmcorvidae]
ah :/
04:52:15 [ianmcorvidae]
well, you have a good night then, thanks for the assistance
04:53:47 [kepstin-laptop]
kepstin-laptop has joined #musicbrainz-devel
04:54:38 [kepstin-laptop]
hmm. I have a box set up with a replicated db already; can I use that to generate a set of database dump files to import into another database?
04:54:57 [kepstin-laptop]
I'd rather not download a new set of dumps if I don't have to :)
04:55:00 [ianmcorvidae]
possibly, though it's worth noting there do exist non-replicated tables
04:55:48 [ianmcorvidae]
admin/ExportAllTables is what does the dumps IIRC
04:56:04 [kepstin-laptop]
well, I can give it a try and see what happens, then.
04:56:38 [kepstin-laptop]
* kepstin-laptop is building a standalone vm on his laptop to allow him to develop riker in places without good internet connections.
04:58:49 [kepstin-laptop]
kepstin-laptop has joined #musicbrainz-devel
05:01:49 [kepstin-laptop]
hmm. can't say that my atom server is very fast at doing a full export :)
05:03:51 [kepstin-laptop]
hmm. /tmp being a 700mb tmpfs isn't really big enough :)
05:04:11 [kepstin-laptop]
* kepstin-laptop tells the script to use a real disc as temp space.
05:05:02 [ianmcorvidae]
* ianmcorvidae meanwhile works on optional test dependencies like phantomjs
05:05:24 [ianmcorvidae]
probably won't bother with selenium though
05:05:57 [kepstin-laptop]
* kepstin-laptop has actually already ordered the parts to upgrade his miniserver to have a fancy dual-core sandy bridge i5 chip.
07:48:46 [warp]
alastairp: t/sql/webservice.sql has a bunch of data, used for the selenium tests, is that what you mean with sample releases?
07:49:31 [warp]
ianmcorvidae: I need to update the phantomjs test. I think our testrunner is broken with current phantomjs releases.
07:49:39 [ianmcorvidae]
oh, bah
07:49:42 [ianmcorvidae]
well, we'll see in a bit
07:50:57 [warp]
(jenkins runs commit 2182d8e54be4cb25fc8fb35801c1f0f0b1c326ca from git://github.com/ariya/phantomjs.git )
07:55:49 [adhawkins]
adhawkins has joined #musicbrainz-devel
07:57:53 [ianmcorvidae]
I'm currently installing current git, which is taking forever to compile
07:58:00 [ianmcorvidae]
so I'll report back after it's done :)
08:15:25 [djce]
djce has joined #musicbrainz-devel
08:21:29 [adhawkins]
ianmcorvidae: Installing where
08:21:30 [adhawkins]
?
08:21:58 [ianmcorvidae]
adhawkins: local machine
08:22:03 [adhawkins]
Ah ok.
08:22:18 [adhawkins]
Have wondered in the past whether I should try to compile an updated version on my debian machine.
08:22:48 [ianmcorvidae]
yeah, no idea there
08:22:52 [ianmcorvidae]
this is an arch linux box :)
08:23:12 [ianmcorvidae]
I'm getting lots of nice warnings about C++11 things :P
08:23:24 [adhawkins]
Right.
08:23:26 [ianmcorvidae]
(well, all the same warning: identifier 'nullptr' is a keyword in C++11)
08:23:38 [adhawkins]
Not sure if there's anything in particular that makes it worth the effort...
08:23:57 [ianmcorvidae]
no idea; there isn't a precompiled package for arch :)
08:24:07 [ianmcorvidae]
(thus my compiling it -- and if I'm going to, may as well use the git version)
08:25:57 [adhawkins]
Ah ok, seems a bit of a surprise that there's no package, although never really used arch.
08:26:33 [ianmcorvidae]
there's a PKGBUILD, you just have to compile the package yourself (there's scripts to do it, etc., and it installs it as a package, it's just not precompiled like debian/ubuntu/etc.)
08:26:44 [ianmcorvidae]
arch is basically half a normal distro and half a from-source distro
08:26:53 [adhawkins]
Ah ok.
08:27:09 [ianmcorvidae]
core stuff that everyone uses they precompile, but more fringe stuff you just get sources and compile (the "Arch User Repository")
08:27:10 [adhawkins]
We use gentoo at work, never fully got to grips with it.
08:27:14 [ianmcorvidae]
yeah
08:27:40 [adhawkins]
Mind you, I started out on SLS and Slackware when there weren't really any packages :)
08:28:01 [adhawkins]
First kernel I ever compiled was 0.99pl12 if memory serves :)
08:28:18 [ianmcorvidae]
heh
08:28:27 [ianmcorvidae]
I probably wasn't using a computer at that point, at least not regularly XD;
08:29:19 [adhawkins]
I'm almost certainly a little older than you :)
08:30:11 [ianmcorvidae]
yup
08:30:28 [ianmcorvidae]
yeah, looks like slackware using 0.99pl12 was late 1993
08:30:54 [ianmcorvidae]
so I guess I might have been starting to use my parents' IBM 486 by then (I would have been four)
08:31:31 [ianmcorvidae]
now I feel like I should go find something to install that on, just for fun
08:34:04 [adhawkins]
Heady days.
08:34:19 [adhawkins]
1993 sounds about right. First used it in Uni, left in 1994.
08:34:53 [adhawkins]
Remember well multiple reinstalls from the (I think) 40 floppies because I'd screwed up a libc or kernel installation.
08:35:12 [ianmcorvidae]
hah
08:35:22 [ianmcorvidae]
yeah, that's the sort of thing I like precompiled packages for :)
08:36:42 [adhawkins]
Yeah, not sure I'd bother with a libc compile these days. I still roll my own kernels though.
08:36:53 [adhawkins]
Just now I know how to install them properly and have multiple boot options :)
08:36:59 [ianmcorvidae]
hehe :)
08:37:11 [ianmcorvidae]
I build a kernel (from a package, but with some patches) for my netbook, to optimize it
08:37:15 [ianmcorvidae]
but I just use the stock one for this computer
08:37:25 [ianmcorvidae]
used to use a digital audio workstation-optimized one but then the package disappered
08:38:48 [adhawkins]
I tend to compile latest kernels for Debian because they're always so far behind.
08:39:03 [adhawkins]
Haven't tried to get 3.x up and running yet though, must do that at some point.
08:40:03 [ianmcorvidae]
ah, yeah
08:40:12 [ianmcorvidae]
arch is fairly bleeding-edge, that's part of why I switched to it
08:40:16 [ianmcorvidae]
(from ubuntu)
08:40:38 [ianmcorvidae]
(the rest was that I was customizing so much that I was barely using ubuntu anymore, so a more stripped-down distro seemed wise :P)
08:40:40 [adhawkins]
My debian box at home is my gateway to the internet primarily, so stability is my main concern :)
08:40:46 [ianmcorvidae]
ah, yeah
08:40:49 [ianmcorvidae]
my VPS runs debian
08:41:05 [ianmcorvidae]
(for the same reason)
08:41:45 [warp]
adhawkins: I did a linux from scratch once (before http://www.linuxfromscratch.org existed)
08:42:29 [warp]
(after messing up a slackware with a libc upgrade attempt that seemed like a good way to learn more about things :)
08:42:50 [warp]
now I just use debian or ubuntu with as little as possible compiled or installed manually.
08:43:05 [adhawkins]
warp: Erk. LFS is probably more hassle than I could be bothered with!
08:43:36 [adhawkins]
All my installs are various flavours of Debian. Main machine is stable, I have VMs for testing and unstable for testing package and compiles with.
08:43:37 [warp]
I was a student, I had lots of time :)
08:46:09 [adhawkins]
:)
08:46:11 [ianmcorvidae]
* ianmcorvidae running phantomjs now, as an aside, we'll see what it does
08:46:15 [ianmcorvidae]
mostly seems to be taking a long time
08:50:07 [ianmcorvidae]
(to do nothing, it seems, unless phantomjs is doing something I just can't see :/)
08:53:23 [warp]
ianmcorvidae: running qunit.t should be instant.
08:53:32 [ianmcorvidae]
heh, okay
08:53:39 [ianmcorvidae]
then I guess new phantomjs doesn't work :)
08:53:44 [warp]
yes. I know :)
08:56:29 [ianmcorvidae]
oh well
08:56:37 [ianmcorvidae]
* ianmcorvidae pushes the one test fix I did find for my codereview
09:04:44 [ocharles]
morning!
09:04:48 [ocharles]
warp: lets to be releasing today?
09:06:32 [warp]
oh yay!
09:07:22 [warp]
ocharles: when do you want to do that?
09:14:07 [ocharles]
warp: just catching up on emails now, shall we say in 45 minutes?
09:14:17 [djce]
djce has joined #musicbrainz-devel
09:19:50 [warp]
sure
09:36:06 [ocharles]
warp: what's an idiomatic way to add a big blob of documentation to a python file?
09:36:07 [ocharles]
""" ?
09:38:10 [warp]
ocharles: a big blob?
09:38:17 [warp]
what is it documenting?
09:38:45 [ocharles]
http://codereview.musicbrainz.org/r/1966/
09:39:03 [ocharles]
I've just stuck """explain""" before each def for now
09:39:11 [ocharles]
I think that's correct. or maybe it's meant to be inside the def
09:39:54 [warp]
inside the def associates the documentation with the def, that's probably what you want.
09:40:03 [ocharles]
ok
09:41:02 [warp]
is the branch not pushed?
09:41:15 [ocharles]
maybe not, i've just pushed it now with my docs though
09:44:22 [warp]
so now, within the python interpreter you can do this:
09:44:24 [warp]
import fabfile
09:44:32 [warp]
help (fabfile.socket_deploy)
09:44:36 [warp]
and get that documentation.
09:45:19 [warp]
(it is available as a string at fabfile.socket_deploy.__doc__)
09:53:04 [ocharles]
ok, cool
09:53:07 [ocharles]
can i ship this now? :)
09:53:20 [ocharles]
I'm really not the right person to be writing documentation, but I more need to see what other people want to do with it
09:56:04 [ocharles]
warp: ok, beta and master have no differences, so I think we can release
09:56:24 [ocharles]
jenkins is happy
09:58:28 [warp]
cool
09:58:47 [warp]
did we actually get any fixes in before the freeze? :)
09:58:48 [ocharles]
i would like to deploy with fabric so I need a ship it there though
09:58:55 [ocharles]
we have 15 things to release apparently
09:58:58 [warp]
wow
09:58:59 [warp]
cool
09:59:02 [warp]
alright
09:59:19 [warp]
so, for me, nothing has changed yet, right?
09:59:33 [ocharles]
what do you mean? as in what you need to do?
09:59:48 [warp]
yeah, I just go into carl or lenny and switch servers off.
09:59:58 [warp]
(well, routing traffic to the servers)
10:00:00 [ocharles]
yea
10:00:03 [ocharles]
and you give me this shipit :P
10:00:10 [warp]
oh right
10:00:54 [warp]
you didn't update the review, so I figured you weren't done
10:01:25 [ocharles]
oh, oops
10:02:09 [ocharles]
done
10:02:58 [warp]
bye bye soundunwound
10:03:02 [ocharles]
* ocharles waves
10:03:59 [warp]
any reason fabfile.py is in the root?
10:04:50 [ocharles]
cause that's where you do deployments from
10:04:54 [ocharles]
and that's what fabric looks for
10:05:04 [warp]
ok
10:05:29 [warp]
This requires you have a 'beta' alias in your .ssh/config file.
10:05:38 [warp]
ocharles: under def test(), is that correct?
10:05:42 [ocharles]
yes
10:05:43 [ocharles]
oh
10:05:46 [ocharles]
that should be 'test'
10:06:01 [warp]
yay, there is value in code reviews.
10:06:13 [ocharles]
:)
10:06:16 [ocharles]
fixed
10:07:10 [warp]
alright, seems fine.
10:07:32 [ocharles]
goody
10:08:42 [warp]
any idea how to check which of the load balancers is currently active? :)
10:08:56 [ocharles]
check their load is what I do
10:09:12 [ocharles]
http://stats.musicbrainz.org/mrtg/drraw/drraw.cgi?Mode=view;Dashboard=1218378978.7925
10:09:24 [ocharles]
looks like carl
10:09:54 [MBJenkins]
Project musicbrainz-server_master build #285: SUCCESS in 13 min: http://ci.musicbrainz.org/job/musicbrainz-server_master/285/
10:09:54 [MBJenkins]
oliver.g.charles: MBS-4899: Remove SoundUnwound from the sidebar
10:09:59 [warp]
* warp nods.
10:10:46 [warp]
ok, I'm ready to take stuff out.
10:11:09 [ocharles]
let me just kick my vpn
10:11:26 [ocharles]
ok, all happy here
10:11:36 [warp]
asterix should be out.
10:12:38 [ocharles]
haha
10:12:39 [ocharles]
catch 22
10:12:54 [ocharles]
to do the update it requires the hosts to have a 'admin/production-deploy.sh'
10:12:58 [ocharles]
but that only comes via git pull
10:13:02 [ocharles]
but that script *does* the git pull
10:13:02 [warp]
:D
10:13:04 [ocharles]
* ocharles changes that
10:13:21 [warp]
well, that's only the first time, right?
10:13:32 [ocharles]
sort of
10:13:43 [ocharles]
if we change it, the deployment will still sort of be a bit wrong
10:13:47 [ocharles]
cause it'll deploy with the old script
10:13:54 [ocharles]
i mean the real solution is to get all this logic into fabfile.py
10:14:08 [warp]
it means any change the production-delpoy.sh needs two releases. first is to update the file. then whatever stuff needs the update.
10:14:23 [ocharles]
well, it won't in a moment
10:14:27 [warp]
ok :)
10:14:28 [ocharles]
i'm changing the fabric file to do the git pull
10:14:33 [ocharles]
no work arounds anymore! :)
10:16:01 [reosarevok]
reosarevok has joined #musicbrainz-devel
10:17:55 [ocharles]
ok, seems to have worked, just doing one more to double check (and I pushed a minor tweak so I want all hosts on the same sha at the end)
10:18:56 [ocharles]
warp: https://gist.github.com/d6a5817fbb196c0084d8 lookin' good!
10:19:00 [ocharles]
warp: asterix done
10:19:07 [warp]
ok
10:19:23 [warp]
pingu is out.
10:21:29 [ocharles]
hmm, compile resources doesn't exit with an error status code if it fails
10:22:18 [ocharles]
warp: ok, pingu done
10:22:46 [warp]
ping back, astro out.
10:23:35 [ocharles]
ah joy, astro can't do a fast forward merge
10:24:06 [warp]
why not?
10:24:11 [ocharles]
dunno
10:24:24 [ocharles]
it had diverged apparently, so I've just splat it with a hard reset
10:24:35 [warp]
yay!
10:24:44 [ocharles]
warp: ok, astro back
10:24:53 [ocharles]
some minor teething problems there, but that was generally a breeze
10:24:58 [warp]
ok, all back.
10:24:58 [ocharles]
you can try it next time ):
10:25:01 [warp]
:)
10:25:01 [ocharles]
:)*
10:25:44 [warp]
* warp tries.
10:25:54 [warp]
(well, not the deploy. I mean the website)
10:26:13 [ocharles]
ha, we have a release on our birthday
10:26:23 [ocharles]
maybe suggest we don't do it on the 23rd :)
10:26:49 [MBJenkins]
Project musicbrainz-server_master build #286: SUCCESS in 10 min: http://ci.musicbrainz.org/job/musicbrainz-server_master/286/
10:26:50 [MBJenkins]
* oliver.g.charles: Add a Fabric deployment script for beta.musicbrainz.org and test.musicbrainz.org
10:26:50 [MBJenkins]
* oliver.g.charles: Add Fabric deployment for our production servers
10:26:51 [MBJenkins]
* oliver.g.charles: Move socket-deploy.sh to admin/
10:26:51 [MBJenkins]
* oliver.g.charles: Fix the test for no local changes
10:26:52 [MBJenkins]
* oliver.g.charles: Fix deploy scripts now they are in admin/
10:26:52 [warp]
haha
10:26:52 [MBJenkins]
* oliver.g.charles: Don't do git diff twice
10:26:53 [MBJenkins]
* oliver.g.charles: Add rudimentary documentation
10:26:53 [MBJenkins]
* oliver.g.charles: Correct documentation for updating test
10:30:25 [ocharles]
warp: so now we hard reset beta and next, right?
10:31:07 [warp]
yes!
10:37:16 [MBJenkins]
Project musicbrainz-server_master build #287: SUCCESS in 10 min: http://ci.musicbrainz.org/job/musicbrainz-server_master/287/
10:37:16 [MBJenkins]
* oliver.g.charles: Minor changes to deployment process
10:37:17 [MBJenkins]
* oliver.g.charles: Remove git-pull from production-deploy.sh as Fabric now does this
10:54:26 [ocharles]
meh, I don't like this 'have to merge 3 times to ship' thing
10:54:36 [ocharles]
but lets see how it goes I guess
11:14:05 [kurtjx]
kurtjx has joined #musicbrainz-devel
11:40:21 [Prophet5]
Prophet5 has joined #musicbrainz-devel
11:41:48 [plaintext]
plaintext has joined #musicbrainz-devel
11:42:54 [ocharles]
warp: we should probably hotfix release http://tickets.musicbrainz.org/browse/MBS-4783
11:42:59 [ocharles]
when I've actually fixed it :)
11:47:06 [djce]
djce has joined #musicbrainz-devel
11:51:54 [warp]
ocharles: oh, meh
11:52:26 [ocharles]
warp: ok, one hotfix, ready to go. jenkins is just running
11:52:28 [ocharles]
maybe you want to do this one?
11:52:32 [ocharles]
apt-get install fabric
11:53:21 [warp]
ok
11:53:28 [warp]
then I need to setup those aliases you mentioned.
11:54:55 [warp]
* warp does not have a ~/.ssh/config file.
11:58:24 [ocharles]
you want me to paste you mine?
12:00:40 [warp]
if you think that will help :)
12:01:38 [MBJenkins]
Project musicbrainz-server_master build #288: SUCCESS in 10 min: http://ci.musicbrainz.org/job/musicbrainz-server_master/288/
12:01:38 [MBJenkins]
* oliver.g.charles: Add the 'reset_test_branches' command to fabfile.py to do post-deployment steps
12:01:39 [MBJenkins]
* oliver.g.charles: Do a socket_deploy after reset_test_branches
12:01:39 [MBJenkins]
* oliver.g.charles: Change socket_deploy to cd into the ~/musicbrainz-server directory
12:01:40 [MBJenkins]
* oliver.g.charles: MBS-4783: Correctly load ISWC objects in RemoveISWC edits
12:02:03 [ocharles]
warp: https://gist.github.com/524beda7bb484fc36aa6
12:02:42 [warp]
right, that's not going to help just yet.
12:03:19 [ocharles]
i see
12:03:52 [ijabz]
ijabz has joined #musicbrainz-devel
12:04:57 [warp]
(as I never use the VPN, I don't have that set up anywhere)
12:06:23 [ocharles]
ok, i don't think you can do this without the vpn
12:07:37 [warp]
I'm pretty sure I can, but I never spent the time figuring out how.
12:18:40 [ocharles]
warp: you getting anywhere?
12:18:44 [warp]
yes, almost done.
12:18:55 [ocharles]
yay
12:20:34 [warp]
ocharles: https://gist.github.com/2988266
12:20:56 [ocharles]
cool :)
12:21:03 [ocharles]
so, shall I drive carl?
12:21:45 [warp]
sure
12:21:58 [warp]
ok, I should be ready now.
12:22:00 [ocharles]
warp: asterix out
12:22:12 [ocharles]
'fab production' and follow the prompts
12:22:52 [warp]
Fatal error: Name lookup failed for asterix
12:23:01 [ocharles]
and 'ssh asterix' works?
12:23:08 [warp]
yes
12:24:46 [ocharles]
try: echo "use_ssh_config = True" >> ~/.fabricrc
12:24:53 [adhawkins]
Are you really going to tag this release? :)
12:24:57 [adhawkins]
* adhawkins is kidding, really.
12:25:19 [ocharles]
adhawkins: sadly it's not up to me
12:25:28 [ocharles]
i noticed that the last release is also untagged
12:25:30 [adhawkins]
Yeah I know, that's why I'm just kidding.
12:25:37 [adhawkins]
Hence my reason for the 'reminder' ;)
12:25:49 [adhawkins]
I can just update based on that commit if necessary.
12:25:59 [adhawkins]
I only use my local server for /ws/2 anyway
12:26:14 [warp]
ocharles: that doesn't seem to make any difference.
12:26:57 [ocharles]
what about env.use_ssh_config = True in the fabfile.py, at the top?
12:27:06 [ocharles]
oh wait
12:27:11 [ocharles]
fabric doesn't actually use ssh
12:27:14 [ocharles]
it uses its own implementation
12:27:17 [ocharles]
so that proxy stuff won't work
12:27:23 [warp]
doh
12:27:40 [ocharles]
you can use -h to specify a host, but that still won't help you
12:28:24 [ocharles]
i'll take care of this release cause it's lunch time now
12:28:27 [ocharles]
you should get the vpn set up :)
12:29:28 [warp]
or fix fabric which is obviously broken
12:29:28 [warp]
:)
12:29:37 [warp]
shall I login to carl then?
12:30:00 [ocharles]
already done
12:30:06 [ocharles]
i was on carl for you :)
12:30:30 [warp]
* warp is confused.
12:30:41 [ocharles]
i was going to drive carl, you were going to drive fabric
12:30:44 [warp]
yes.
12:32:01 [warp]
since i'm not driving fabric, I figured we swap back and I drive carl again.
12:32:14 [ocharles]
ah, well I just carried on
12:32:17 [ocharles]
I mostly want to eat lunch :)
12:32:21 [warp]
ok
12:32:22 [ocharles]
anyway, hotfix pushed
12:32:26 [ocharles]
so i'm going to go eat, bbiab
12:32:30 [warp]
great, thanks
12:32:33 [warp]
enjy lunch!
12:32:36 [warp]
+o
12:51:37 [plaintext]
hey guys!
12:51:55 [plaintext]
how would I go about accessing/modifying the db in python on a VM?
12:52:11 [plaintext]
I found a PyGreSQL module for python, but it seems it's not installed on the VM
12:53:22 [ocharles]
plaintext: psycopg2 is one of the go-to python modules for db access
12:53:30 [plaintext]
thanks!
12:53:34 [ocharles]
but you'll need to install whatever library you want on the VM
12:53:38 [ocharles]
should be doable with apt-get
12:53:59 [ocharles]
check 'man apt-get' and 'man apt-cache' -- i don't use ubuntu so i'm not much help
12:54:24 [plaintext]
oh so it's not installed yet
12:54:32 [ocharles]
probably not
12:54:37 [plaintext]
I see, thank you
13:03:58 [djce]
djce has joined #musicbrainz-devel
13:08:27 [kurtjx]
kurtjx has joined #musicbrainz-devel
13:11:19 [warp]
plaintext: ofcourse you can also change the permissions on the VM to allow access to the postgres database from your host or another machine in the network.
13:11:48 [plaintext]
warp: thanks, but I'm not sure why I would need that
13:21:22 [djce]
djce has joined #musicbrainz-devel
13:25:37 [plaintext]
I'm quite new to this :)
13:29:40 [ocharles]
plaintext: it means you can use the VM as just a database server and do your development on the rest of your machine
13:29:50 [plaintext]
oh I see
13:47:23 [alastairp]
warp: right. that looks like it would be enough. is that imported for all tests?
13:50:33 [plaintext]
I can't seem to find where the connection details are stored
13:50:39 [plaintext]
Could anybody help?
13:51:17 [ocharles]
plaintext: lib/DBDefs.pm
13:51:21 [alastairp]
/etc/postgresql/9.1/main/pg_hba.conf
13:51:23 [alastairp]
oh
13:51:28 [plaintext]
thanks!
13:51:33 [alastairp]
configuration, not postgres
13:53:32 [ocharles]
warp: http://tickets.musicbrainz.org/browse/MBS-2301 is now overdue, but I don't feel comfortable doing it. Can I assign it to you?
13:59:55 [warp]
sure
14:20:05 [thejav]
thejav has joined #musicbrainz-devel
14:22:55 [thejav]
hello, I'm working on an app that shows discographies for several artists. I really want to minimize my queries to MusicBrainz, and so far I have a query that gets me exactly what I need in one call, save for release dates. It's this: http://www.musicbrainz.org/ws/2/release-group/?type=artist&query=artist:"artist name"+type:album+primaryType:album+status:official Is it possible to get releaseDates in that response?
14:29:03 [thejav]
or, to restate, it would be nice if release-group would give some details about the various release dates contained by it's releases without having to do another lookup
14:33:47 [kepstin-laptop]
yay, musicbrainz devel server set up on my laptop.
14:35:32 [kepstin-laptop]
hmm. the release group should include the 'earliest release date' field, but that might not work in query mode.
14:35:56 [kepstin-laptop]
of course, you really should be doing browse requests by artist mbid, rather than searches by artist name...
14:36:39 [kepstin-laptop]
if there are two artists with the same name, your query will get the release groups from both of them.
14:41:33 [thejav]
it's not perfect, but my goal is to take the name of an artist and do one query to get their discography, including release dates. If I change that query to 'release' instead of 'release-group', i have pretty much what I want if I group the results by release-group. it does have the problem you've mentioned
14:42:04 [thejav]
what is the catalog number of a label, by the way?
14:42:17 [kepstin-laptop]
if you do a browse request by artist mbid, you should get an earliest release date field in the release group element.
14:42:33 [Mineo]
there is no catalog number of a label, there's only the label code
14:43:11 [kepstin-laptop]
the catalog number is basically a number that a store would use to order a copy of a release out of a label's catalog.
14:43:32 [kepstin-laptop]
just a helpful id number for releases.
14:43:59 [kepstin-laptop]
(a lot of modern labels just use the barcode number, though)
14:44:45 [ruaok]
ruaok has joined #musicbrainz-devel
14:45:26 [thejav]
ah I see.
14:47:17 [kepstin-laptop]
hmm. if you do a browse request by mbid, you can't use the search server to filter by release/release group status and type.
14:48:41 [kepstin-laptop]
so e.g. http://musicbrainz.org/ws/2/release-group?artist=b6c18308-82c7-4ec1-a42d-e8488bce6618 gives you release dates, but you'd have to filter the types on your end, and you can't even see the statuses
14:51:25 [kepstin-laptop]
thejav: you should probably open a ticket on tickets.musicbrainz.org asking for the 'first-release-date' field to be added to the index search results for release-groups.
14:53:29 [thejav]
yeah statuses are a deal breaker. but I can work with using releases instead of release groups. in your opinion, should I take the earliest release date for a release-group and call that the "release date" for a particular album? i don't want to vary it by country, i just want something consistent
14:55:00 [kepstin-laptop]
the earliest release date should be generally 'close enough'. Depending on what you're using it for, you might want to use just the year+month and ignore the day.
14:56:54 [alastairp]
ocharles: I have a branch that I've made some tests on
14:57:04 [alastairp]
should I be merging somewhere and pushing so that tests can run?
14:58:49 [ocharles]
no, tests only run against master atm
15:06:09 [hawke_1]
hawke_1 has joined #musicbrainz-devel
15:09:49 [alastairp]
ok. so I don't merge into there until it's had a ship-ip?
15:09:54 [alastairp]
ship-it
15:11:38 [ocharles]
you don't merge there at all :)
15:11:53 [ocharles]
only a handful of people have rights to merge to master
15:12:00 [alastairp]
ok
15:12:02 [ocharles]
all you need to worry about is getting a ship it and having a branch public
15:12:10 [ocharles]
i'll take care of integrating
15:12:12 [alastairp]
ah, fine
15:12:30 [alastairp]
ok, next question
15:12:44 [alastairp]
getting fixtures up for tests
15:12:52 [alastairp]
can I use the data models?
15:12:59 [alastairp]
or is it preferred to insert sql?
15:13:17 [ocharles]
a slight preference for sql
15:13:24 [ocharles]
as that's generally what the other tests do
15:13:30 [alastairp]
ok
15:14:19 [alastairp]
when is stuff from t/sql imported? always? do I need to kick it off?
15:14:27 [ocharles]
nope, you have to do that manually
15:14:43 [ocharles]
if you do MusicBrainz::Server::Test->prepare_test_database($c, '+foo'), t/sql/foo.sql will be loaded
15:14:53 [ocharles]
which is admitted, slightlly cryptic
15:14:54 [alastairp]
great
15:15:14 [ocharles]
i am personally moving towards inlining SQL in the test methods themselves with $c->sql->do(<<'EOSQL');
15:15:21 [ocharles]
but it remains to be seen if that's actually a good idea or not
15:15:46 [alastairp]
right, but surely that depends on the size of the fixture needed?
15:15:58 [ocharles]
even bigger ones I tend to inline these days
15:18:25 [warp]
ocharles: I don't like the heredocs, but that's mainly because of lousy editor support.
15:18:38 [ocharles]
warp: M-x sql-mode usually stops my grumbling :)
15:18:56 [ocharles]
but i agree, it's not perfect
15:19:10 [ocharles]
but with separate fixtures they just seem to grow and grow and there's all sorts of bizzare dependencies
15:19:15 [warp]
ocharles: I think it messes with indenting. (so I have to be careful with M-x indent-region)
15:19:55 [alastairp]
ocharles: right, but because these tests are usually full end-to-end you often need significant fixtures
15:20:02 [alastairp]
to get to the bit that you want to test
15:20:20 [ocharles]
alastairp: what do you mean by full end-to-end?
15:20:35 [ocharles]
we have some sort of 'legacy' style tests which have 'test all' which are bad examples
15:20:46 [ocharles]
Edit::Work::RemoveISWC is an example of a more modern test
15:20:47 [alastairp]
the fact that your tests spin up a database + webserver + make a request
15:21:28 [ocharles]
yea, you do need a fair bit of preliminary data to satisfy FKs
15:23:47 [alastairp]
hmm, but yeah. having common fixtures means that adding something to it for test x might break test y
15:24:18 [ocharles]
yep, and that's happened
15:24:47 [ocharles]
the current method means you need to be pretty handy with sed when we do schema change releases, but thankfuly they don't come around that often
15:25:20 [alastairp]
do we still push branches to git.mb?
15:25:27 [alastairp]
or has it moved to gh
15:26:14 [ocharles]
nope, still git.mb
15:26:28 [alastairp]
ah. what's the ssh url?
15:27:10 [ocharles]
ssh://mbgitrw@git.musicbrainz.org:10015/musicbrainz-server.git
15:27:31 [alastairp]
and, lastly,
15:27:40 [alastairp]
should I merge origin/master into my branch before I push?
15:27:53 [alastairp]
(ff, nonff always?)
15:27:57 [ocharles]
not essentially, but always nice to know there's no merge conflicts
15:28:00 [alastairp]
ok
15:28:08 [ocharles]
you won't be able to do a ff merge, assuming you've actually done work ;)
15:28:17 [alastairp]
true :)
15:28:20 [ocharles]
rebasing on top of master could be good though
15:28:23 [ocharles]
if it's your first push
15:28:30 [ocharles]
but again, not essential
15:28:42 [alastairp]
ok. How about I don't do that now, and we can move upto it
15:28:54 [ocharles]
yea, that's fine
15:29:04 [ocharles]
the only time I tend to make people merge master into their own branch is when it conflicts
15:41:51 [warp]
* warp nods.
15:44:24 [warp]
gah!
15:45:09 [warp]
I was wondering why my ajax upload test wasn't working. fake s3 was running, all the other caa bits are running, but for some reason it couldn't connect to localhost:5050
15:45:26 [warp]
this is to test internet explorer support.
15:46:13 [warp]
(obviously that internet explorer is running on a different machine, so localhost should've been an actual host. *facepalm*)
15:46:19 [ocharles]
:)
15:51:18 [Leftmost]
Leftmost has joined #musicbrainz-devel
15:51:18 [Leftmost]
Leftmost has joined #musicbrainz-devel
15:51:25 [ocharles]
oh, alastairp, when you create reviews, fill in the 'branch' and 'bug' fields
15:51:38 [ocharles]
no need to change the one that already has a shipit, but just a note for going forward
15:54:53 [alastairp]
oh, there's branch and bug fields? great!
15:55:19 [ocharles]
yep, i'll be making use of them when I get round to automating more code review/shipping
15:56:46 [hawke_2]
hawke_2 has joined #musicbrainz-devel
16:10:35 [alastairp]
does around run_test => sub {} run before each test, or before all tests?
16:11:22 [ocharles]
before each test
16:11:32 [ocharles]
sub BUILD and sub DESTROY run around the entire class
16:15:15 [alastairp]
ok, cool
16:34:16 [plaintext]
I pushed new code to the repository
16:34:21 [plaintext]
Created the diff
16:34:31 [plaintext]
and when I'm trying to add the diff on codereview I get this:
16:34:37 [plaintext]
"The file "log_sanitizer.py" (revision ff118454582bbe54587eea65e6ac284889e47981) was not found in the repository"
16:34:44 [plaintext]
Does anyone know what's wrong?
16:34:49 [reosarevok]
plaintext: did you do a git add?
16:34:58 [reosarevok]
(to add the new file to the list of commited files)
16:35:14 [reosarevok]
Oh, wait, on codereview :)
16:35:41 [reosarevok]
huh, never seen that error myself :)
16:35:44 [reosarevok]
ocharles?
16:36:52 [demosdemon]
code-review needs a reference to the original file before the patch to display the diff correctly… usually post-review does this automatically
16:38:08 [reosarevok]
huh
16:38:11 [alastairp]
I got that when I did my diff backwards
16:38:17 [reosarevok]
I've repeatedly added new files and never got an issue
16:38:22 [reosarevok]
So maybe it's what alastairp says :)
16:38:27 [ocharles]
sounds possible
16:38:36 [ocharles]
a) how are you creating the review and b) which repository are you using
16:39:22 [demosdemon]
he saw none of that :P
16:40:04 [ocharles]
oh, I don't have parts or joins visible
16:40:23 [demosdemon]
he just pinged out
16:42:34 [reosarevok]
ruaok or ocharles: it seems replication has been broken for a few hours
16:42:44 [ocharles]
yes
16:42:46 [ocharles]
cron is broke
16:42:48 [ocharles]
working on that
16:43:00 [ruaok]
oh, its on our end?
16:43:05 [ocharles]
yea
16:43:07 [ruaok]
what's broke?
16:43:26 [ocharles]
cronjobs try and run carton themselves, but carton isn't on cron $PATH
16:43:51 [ocharles]
I had this problem with beta and fixed it there, but astro isn't playing nicely
16:44:39 [ruaok]
can you please look into fixing this asap?
16:44:57 [ocharles]
yea, just gave up trying to fix this with cron and stuck an export in the hourly.sh file
16:45:40 [ocharles]
hopefully they'll run this hour
16:45:47 [ruaok]
k
16:51:47 [alastairp]
hm
16:51:53 [alastairp]
ocharles: I may have screwed up that push
16:52:02 [ocharles]
alastairp: no worries, just nuke it with a --force
16:52:05 [alastairp]
wrong branch. one sec
16:52:08 [ocharles]
i haven't pulled
16:52:22 [ocharles]
story of my life right there
16:52:22 [ocharles]
:P
16:52:44 [alastairp]
yeah. I shouldn't work on 3 separate branches at the same time
16:52:48 [alastairp]
and jump between them
16:53:21 [ocharles]
no apparently you should be working on...
16:53:22 [ocharles]
git branch | wc -l
16:53:23 [ocharles]
75
16:53:28 [alastairp]
hah
16:53:31 [alastairp]
trie
16:53:32 [alastairp]
true
16:53:43 [alastairp]
nah, I got it. I was in the wrong ssh session
16:53:54 [alastairp]
I need to set up git-sh I think
16:54:12 [ocharles]
let me know when you're happy and i'll get your work on test
16:54:20 [ocharles]
beta in fact, as I gave you a ship it
16:58:46 [alastairp]
turns out it wasn't an issue. it's good
17:08:04 [ocharles]
ruaok: do you know how long hourly.sh normally takes?
17:08:16 [alastairp]
hopefully less than an hour
17:08:21 [ocharles]
:)
17:08:28 [ruaok]
normally 8 minutes or so.
17:08:40 [ruaok]
but if its not run for hours, then I have no idea.
17:08:50 [ocharles]
right, it'll take a little longer
17:08:56 [warp]
X * 8 minutes?
17:09:15 [ocharles]
doubt it
17:09:29 [warp]
I hope it's not more :)
17:09:30 [ocharles]
8 minutes + (X - 1) * Y more likely
17:12:09 [ocharles]
yay, it worked
17:12:50 [ocharles]
* ocharles removes the bit that makes bash warn about job control
17:13:43 [ocharles]
ruaok: I assume it takes a while for the replication packet to show up in ftp?
17:14:08 [alastairp]
if I want to test how many columns in a table, do I use xpath to count it, or mechanize to get the table and assert?
17:14:11 [ruaok]
yes, but they will show up on data right after the sync
17:14:37 [ocharles]
alastairp: you use mechanize to get the page and then work xpath magic on it
17:14:43 [ocharles]
but why are you interested in how many columns their are?
17:14:45 [ocharles]
there*
17:14:59 [alastairp]
because the number of columns changes depending on if you're logged in or not
17:15:06 [ocharles]
ah
17:15:16 [ocharles]
might be easier to check if the data inside that column is or isn't present
17:15:29 [ocharles]
that's probably what's more interesting to test
17:15:38 [ocharles]
ok, time to get myself home
17:15:49 [reosarevok]
ocharles: one hour for replicated servers to update then? (waiting for my changes and adds to appear in http://mb.lmfao.org.uk/labels/rv so I can keep working on it :) )
17:17:10 [ruaok]
the packet is out.
17:17:16 [ruaok]
clients should be able to fetch it now
17:17:32 [ocharles]
great
17:17:38 [reosarevok]
Cool :D
17:17:41 [ocharles]
back in 45
18:25:38 [the_metalgamer]
the_metalgamer has joined #musicbrainz-devel
18:32:15 [adhawkins]
adhawkins has joined #musicbrainz-devel
18:46:00 [warp]
warp has changed the topic to: http://musicbrainz.org/#devel | agenda: SlowDown (warp), browser miminum requirements (warp)
18:46:55 [ocharles]
ocharles has changed the topic to: http://musicbrainz.org/#devel | agenda: Reviews, SlowDown (warp), browser miminum requirements (warp)
18:59:39 [ruaok]
ruaok has changed the topic to: agenda: Reviews, SlowDown (warp), browser miminum requirements (warp), soundunwound (ruaok)
19:00:04 [plaintext]
plaintext has joined #musicbrainz-devel
19:00:05 [warp]
hello!
19:00:11 [ruaok]
<BANG>
19:00:12 [alastairp]
hi
19:00:13 [plaintext]
uh sorry
19:00:13 [ruaok]
meeting time!
19:00:25 [ruaok]
plaintext: sorry for what? your timing is perfect?
19:00:30 [ruaok]
warp: you go first for review!
19:00:34 [warp]
eep!
19:00:52 [plaintext]
I'll explain later :)
19:01:14 [warp]
I had some fun trying to review ianmcorvidae's split-domains branch.
19:01:39 [warp]
this was the first I was trying to run the i18n code, so I hit a whole bunch of issues which I spent some time debugging with ianmcorvidae
19:01:44 [warp]
+time
19:02:54 [warp]
I spent some time trying to figure out whether the internet explore XDomainRequest object would allow file uploads, so I can add IE support to the fancy multifile upload.
19:03:01 [warp]
(it doesn't)
19:03:48 [warp]
I ported my fake s3 php file to perl (with plack), because oliver and ian don't like PHP :)
19:03:53 [ocharles]
:)
19:04:02 [ocharles]
good learning experience too!
19:04:12 [ianmcorvidae]
I don't like yet-more dependencies :P
19:04:29 [warp]
I think that's on code review, if it isn't it should be.
19:04:57 [ocharles]
it is
19:05:03 [warp]
and I did some debugging of thumbnail support at the internet archive with nikki
19:05:21 [ruaok]
I've passed the request for those types over to the archive.
19:05:26 [ruaok]
thanks for making that bucket
19:05:31 [warp]
this also allowed me to test the ajax upload on the archive.org servers (I hadn't done that yet), from which I learned that we need more CORS headers.
19:06:01 [warp]
I sent an e-mail to sam about that, he said he would try to implement that.
19:06:17 [warp]
erm, that's about it.
19:06:31 [ocharles]
it sounds like a good week
19:06:34 [ruaok]
sweet. sounds like a busy week.
19:06:38 [ruaok]
ocharles?
19:06:46 [ocharles]
not so busy for me i'm afraid
19:07:03 [ocharles]
I took friday off because i had to catch up with boring finance crap - 4 months behind takes a bit of time to catch up on :(
19:07:21 [warp]
lol
19:07:25 [ocharles]
but monday and tuesday I was working as normal. I spent some time prototyping a way to use Varnish with our web service, and have it heavily cached
19:07:35 [ocharles]
https://github.com/ocharles/varnish-cache-invalidation
19:07:56 [ocharles]
The idea is: the backend servers (musicbrainz) do exactly what they do now, but they also include a 'Dependencies' header
19:08:29 [ruaok]
did access to the splunk logs help?
19:08:30 [ocharles]
which is of the form artist=5038, etc. Now, when artist=5038 changes, I just tell Varnish to forget everything that it cached where Dependencies matches the regular expression '.*artist=5038.*'
19:08:41 [ocharles]
a bit, I didn't spend much time playing with them
19:09:07 [ocharles]
I was mostly doing this because I wasn't sure what to work on instead :) But today I started work on a new database-backed reports system, and am about half way there
19:09:17 [ocharles]
so hopefully tomorrow, I will have some reports filterable by subscriptions
19:09:36 [ruaok]
ocharles: when in doubt on what to work on, work on the new edit system
19:09:51 [ocharles]
right, i haven't started plugging on with that yet actually
19:09:57 [ocharles]
I'll bump that forward
19:10:00 [ruaok]
time to start. :)
19:10:01 [ruaok]
k
19:10:10 [ocharles]
I did say last week I'd look into the 502 stuff a bit more, but I haven't got round to that. Maybe this week, we'll see how it goes
19:10:14 [warp]
ocharles: let me know when there is something for me to work on related to NES
19:10:27 [ocharles]
oh, and today we did our first Fabric deployment. it didn't go entirely smoothly, which was somewhat to be expected
19:10:29 [ruaok]
* ruaok goes next
19:10:39 [ocharles]
but it was still much more fun that typing stuff manually
19:10:42 [ocharles]
ok, all done!
19:10:50 [ruaok]
I should report on my last two weeks, since I was busy flying last week.
19:10:57 [ianmcorvidae]
ocharles: now or later, could you give me a quick rundown of what needs improving before rolling out varnish/cache invalidation/etc? I'd really like that to push forward, and if you're supposed to work on the edit system... :)
19:11:07 [ocharles]
ianmcorvidae: sure thing, grab me post-meeting
19:11:10 [ianmcorvidae]
k
19:11:16 [ruaok]
week, prior was spent in london
19:11:29 [ruaok]
I had a series of good meetings there.
19:11:45 [ruaok]
spotify, 7 digital and west 10 entertainment, universal music.
19:12:09 [ruaok]
7 digital was fun… several engineers there are using their 20% time to examine musicbrainz.
19:12:22 [ruaok]
they had a lot of interesting questions.
19:12:35 [ruaok]
it seems that their curiousity has been piqued by MB.
19:12:48 [ruaok]
then I zipped down to BCN.
19:13:00 [ruaok]
in BCN I presented MB for the music hack day.
19:13:12 [ijabz_]
ijabz_ has joined #musicbrainz-devel
19:13:18 [ruaok]
then later on thursday I sat on a panel as part of the mires conference.
19:13:26 [ruaok]
that panel went well and got more exposure for MB.
19:13:32 [warp]
I met one the 7 digital people a music hack day in amsterdam some time ago, they just started looking into MB at the time. good to see they're still interested and making some progress there :)
19:13:46 [ruaok]
then I hacked on huesound for spotify.
19:14:03 [ruaok]
I won the spotify prize (with one other fellow) for the best spotify based hack.
19:14:25 [ruaok]
then I spent the weekend at sonar (the music festival) and saw new order. :-D
19:14:40 [ruaok]
on monday I flew back and have been trying to get back into the swing of things.
19:14:55 [ruaok]
though I did mostly catch-up stuff and setting plaintext on the right course.
19:15:09 [ruaok]
I have a ton of things to do for catch up -- not so fun.
19:15:24 [ianmcorvidae]
you mentioned summit-related research in BCN -- any update on that?
19:15:38 [ruaok]
and there is a pile of hosting related things waiting for me when I return to SLO.
19:15:42 [ruaok]
ianmcorvidae: yes.
19:16:03 [ruaok]
the MTG (music technology group) at University Pompeu Fabra is quite cool.
19:16:20 [ruaok]
the are the hosts to the comp music project, and I've met the whole comp music team.
19:16:22 [ruaok]
very cool.
19:16:57 [ruaok]
when I mentioned to them that we're looking for a place to hold our next summit, they jumped at the chance and offered factilities for us to host the summit.
19:17:04 [ruaok]
which is really awesome.
19:17:19 [ruaok]
so, the next summit will be in november in Barcelona!
19:17:27 [warp]
woot!
19:17:40 [ruaok]
exact details TBD -- I need to work down a backlog of stuff before I dive into that.
19:17:49 [ruaok]
oh, and its too soon to think about plane tickets.
19:18:04 [ruaok]
we should wait until sept to buy them -- fall fare sales kick in after sept 15.
19:18:11 [ocharles]
cool
19:18:32 [ruaok]
I'll do my best again to get sponsorship for the summit again. see if we can sponsor a few people to come to the summit.
19:18:48 [ruaok]
ok, enough for me.
19:18:57 [ruaok]
who wants to go next?
19:19:16 [ruaok]
gsoc? reosarevok? ianmcorvidae? ijabz?
19:19:19 [ianmcorvidae]
I can go
19:19:35 [ruaok]
go
19:19:38 [ianmcorvidae]
mine's pretty quick, I mostly made a whole bunch of revisions to the codereview that just shipped for i18n stuff
19:19:52 [ianmcorvidae]
found one change that's needed for switching to the new wiki, wikidocs-wise
19:20:07 [ianmcorvidae]
hoping to look at styling for that at some point, though I can get back to i18n work now as well :)
19:20:15 [ianmcorvidae]
and done :)
19:20:18 [ruaok]
nice.
19:20:28 [ruaok]
plaintext? bitmap? alastairp_ ?
19:20:33 [ruaok]
demosdemon?
19:20:37 [alastairp_]
whoops
19:20:42 [bitmap]
I could give an update
19:20:45 [ruaok]
bitmap: go
19:20:49 [alastairp_]
someone turned off my IRC machine. lucky I'm still here
19:20:55 [bitmap]
I have the basic UI working, so adding and changing relationships all works *on the page*
19:21:05 [ruaok]
bitmap: the comp music folks are quite looking forward to the relationship editor, btw.
19:21:09 [bitmap]
this took a very long time, because there were many problems to deal with when a relationship appeared more than once (it has to keep everything in sync). this is solved now
19:21:16 [bitmap]
cool :)
19:21:25 [bitmap]
it's viewable at http://bitmap.mbsandbox.org/ if anyone's interested (click edit relationships on a release page)
19:21:34 [warp]
bitmap: when can we expect stuff on code review?
19:21:50 [bitmap]
I want to get a review out soon, but I think it'll have to wait another week until actual submission/form error handling is working at least basically, which is what I'm doing this week
19:22:13 [bitmap]
I have around 75% of my first three tasks done, so I'm working at a good pace though
19:22:25 [bitmap]
that's all I have
19:22:38 [warp]
:)
19:22:51 [ruaok]
plaintext?
19:22:54 [plaintext]
I can go
19:22:57 [plaintext]
it's a short one
19:23:11 [plaintext]
So I've just arrived and started working again
19:23:20 [plaintext]
I did what ocharles suggested about log sanitizing
19:23:34 [plaintext]
So I parsed URLs and checked the email regex on values
19:23:40 [ruaok]
uh oh. listening to ocharles is mostly a bad idea. :)
19:23:44 [ocharles]
amen
19:23:47 [plaintext]
which gave me a list of entries. 99% was just real data
19:24:00 [Prophet5]
Prophet5 has joined #musicbrainz-devel
19:24:00 [plaintext]
but at least this showed me one more log entry that we didn't sanitize before
19:24:06 [plaintext]
so it was a good idea after all :)
19:24:15 [ocharles]
:)
19:24:18 [ruaok]
plaintext: that was the goal I had hoped for.
19:24:22 [ocharles]
HA. I do have value
19:24:22 [plaintext]
yeah
19:24:27 [plaintext]
Then I started working on creating the table on the VM
19:24:30 [ruaok]
I think we're good to go on the sanitizing script.
19:24:35 [ocharles]
+1
19:24:43 [MBChatLogger]
http://splunk.musicbrainz.org
19:24:43 [plaintext]
but I was unable to reach splunk.mb.org so I'll have to ask about it after the meeting
19:24:56 [plaintext]
I did create the script though and it works on my machine
19:24:59 [alastairp]
alastairp has joined #musicbrainz-devel
19:25:02 [ruaok]
plaintext: can you update the script on codereview so we can have a final look?
19:25:08 [ruaok]
plaintext: ok
19:25:10 [plaintext]
yeah I have a minor problem with that
19:25:19 [plaintext]
I asked here but then I pinged out and I didn't realize
19:25:33 [plaintext]
I've read the logs and alastairp_ suggested that I might have done the diff backwards
19:25:46 [plaintext]
because codereview gave me an interesting error
19:26:05 [plaintext]
I tried but it didn't work so I still get an error when I'm trying to add the diff
19:26:41 [ruaok]
ok, lets go over the error after the meeting.
19:26:44 [plaintext]
sure
19:26:47 [plaintext]
I think that's it
19:26:53 [ruaok]
and the not being able to access splunk.
19:26:55 [ruaok]
k.
19:26:58 [ruaok]
alastairp_: ?
19:27:02 [plaintext]
I'm close to getting it done apart from these small errors
19:27:06 [alastairp]
right
19:27:23 [alastairp]
I wasn't at the meeting last week, but in the last few weeks I've attacked my first tickets again
19:27:35 [alastairp]
I got feedback from ocharles and warp, mostly around "unit tests, pls"
19:27:53 [alastairp]
I've written them and re-submitted. Hopefully those tickets will be merged this week
19:28:01 [ocharles]
i would think they will, yea
19:28:09 [alastairp]
so I'll start work on some more (I"ve blogged which ones)
19:28:35 [alastairp]
I also spent a day a week or so back doing some work on my python bindings
19:28:54 [alastairp]
another week's work and they'll be ready to release the next version, which will include support for all the schema changes
19:29:27 [alastairp]
also, related to ruaok's update, I'm doing a presentation on mb at the next CompMusic workshop
19:29:37 [alastairp]
in Istanbul in a few weeks time
19:29:49 [alastairp]
so I spent some time working on that
19:29:59 [alastairp]
that's me
19:30:04 [ruaok]
oh, I forgot. the comp music team would like to add a few data fields to our DB.
19:30:15 [ruaok]
mostly around india and turkish music.
19:30:25 [ruaok]
I've asked them to come up with a detailed proposal for this.
19:30:33 [ruaok]
so we'll see when that comes about.
19:30:44 [reosarevok]
We can probably talk about that in person in Nov
19:30:45 [ruaok]
ok, anyone else for review?
19:30:50 [alastairp]
I expect I'll be a compmusic/mb liason come oct
19:30:52 [reosarevok]
Unless they're in a big hurry...
19:30:57 [ijabz]
nothing from me
19:31:03 [ruaok]
reosarevok: I think they might shoot for the oct release.
19:31:12 [reosarevok]
hmmm
19:31:15 [ruaok]
alastairp: +1 on you being the liason!
19:31:21 [reosarevok]
Well, if they actually write a detailed proposal this time...
19:31:37 [ianmcorvidae]
demosdemon not around, I take it?
19:31:53 [reosarevok]
* reosarevok seems to remember mentioning more detailed requests and not getting them last time but I guess now they saw we're serious about adding stuff
19:31:55 [reosarevok]
:)
19:32:12 [ruaok]
reosarevok: well, we can ignore them until they come up with a solid propsal.
19:32:23 [ruaok]
I outlined what piece of data we're looking for in their proposal
19:32:28 [reosarevok]
Nah, I mean, I *do* want them to send it :)
19:32:45 [reosarevok]
No need to ignore them, we can help a bit :)
19:32:50 [ruaok]
well, maybe you'll get a chance to meet them in person. :)
19:33:04 [ruaok]
ok, lets move on then.
19:33:09 [ruaok]
warp: slowdown!
19:33:56 [warp]
ok!
19:34:11 [warp]
we've had some SlowDown problems again from the internet archive this weekend.
19:35:10 [warp]
my main question on this topic is if this is something we need to poke the archive about when it happens?
19:35:23 [ruaok]
we should.
19:35:36 [ruaok]
though we might as well set a cron job to do that.
19:35:42 [warp]
and considering it is quite frequent, also ask if there is anything they can improve on their end to er.. get less of them :)
19:35:46 [ruaok]
seems that they break reliably on the weekends
19:36:02 [ruaok]
I'll mention it again to brewster next time we chat
19:36:29 [warp]
somewhat related, I've asked sam to set CORS headers on this response, so we can at least show it to our users in a nicer fashion.
19:37:07 [warp]
ruaok: so, when it breaks, should we report to you, and you'll contact archive.org?
19:37:36 [ruaok]
if I am around. otherwise send mail to sam directly.
19:37:56 [warp]
ok, great, that's what I wanted to know.
19:38:07 [warp]
warp has changed the topic to: agenda: browser miminum requirements (warp), soundunwound (ruaok)
19:38:16 [warp]
shall I continue to the next topic? :)
19:38:20 [ruaok]
yep
19:38:50 [warp]
Now that I'm mucking with cutting edge xmlhttprequests :), I'm again wondering what browsers I should be supporting
19:39:12 [ocharles]
chrome and nothing else!
19:39:21 [warp]
I don't think we've ever made any hard decisions on that topic, have we?
19:40:00 [ocharles]
i don't think so
19:40:31 [ruaok]
nope.
19:40:49 [ruaok]
we have the google analytics stuff that should give us a good clue.
19:41:02 [warp]
who has access to that?
19:41:05 [ruaok]
next time you see navap, ask him to print a PDF of that report.
19:41:06 [warp]
navap?
19:41:14 [navap]
I'm pulling it up now
19:41:16 [warp]
can I get access? :)
19:41:26 [ruaok]
I dont see why not
19:41:55 [warp]
ok, let me look at the data this week then, and propose something at the next meeting?
19:42:00 [ruaok]
perfect.
19:42:10 [ruaok]
done then?
19:42:13 [warp]
navap: can you give me access?
19:42:14 [ruaok]
(with this topic)
19:42:15 [warp]
ruaok: yes.
19:42:19 [ruaok]
k
19:42:21 [navap]
warp: See PM
19:42:31 [ruaok]
the final topic is about soundunwound.
19:42:38 [ruaok]
the site has been wound down.
19:42:48 [warp]
definitely unwound
19:42:53 [ruaok]
it seems that it never got any critical momentum.
19:43:13 [ruaok]
the good part is that amazon would like to continue our relationship.
19:43:23 [ruaok]
and they have started using our data in the amazon store as well.
19:43:28 [ruaok]
it was nice to get that confirmed.
19:43:48 [ruaok]
that was it.
19:43:50 [hawke__]
Heh
19:43:52 [ruaok]
final thing.
19:44:00 [hawke__]
That was pretty evident…
19:44:03 [ruaok]
I would like to see us schedule a meeting to do bug triage.
19:44:21 [ruaok]
not a fleeting 10 minutes here or there, but at least an hour of it.
19:44:42 [ruaok]
ocharles, reosarevok, warp: how about sometime this week, standard meeting time?
19:44:57 [warp]
ruaok: sure.
19:44:59 [reosarevok]
* reosarevok checks agenda
19:45:01 [ruaok]
tomorrow?
19:45:09 [ocharles]
any reason we can't just do 20 minutes some days during the week?
19:45:25 [ruaok]
ocharles: because its hard to convene people.
19:45:31 [ruaok]
and 20 minutes never gets us far.
19:45:38 [ocharles]
but more than 20 minutes people switch off
19:45:42 [ruaok]
I'd love to have a bit of a catch-up session.
19:45:44 [ocharles]
it's hard to focus for 20 minutes as it is
19:45:50 [ocharles]
(on triaging)
19:45:59 [reosarevok]
Tomorrow works for me
19:46:01 [ruaok]
then lets see if we can do half an hour tomorrow.
19:46:05 [reosarevok]
See http://en.wikipedia.org/wiki/UEFA_Euro_2012#Semi-finals for when it does not :p
19:46:09 [ocharles]
half now, half tomorrow?
19:46:24 [warp]
reosarevok: haha
19:46:25 [ruaok]
I can do some.
19:46:28 [ijabz]
can i ask a related question to bug triage
19:46:34 [ruaok]
ijabz: sure
19:46:53 [ijabz]
sometimes bugs on bug list leap getting moved to next release because they don't get fixed for the release ...
19:47:11 [ijabz]
then do they sometimes just drop off the schedule completely ?
19:47:34 [ocharles]
if that's happening, it usually means we need to discuss the ticket
19:47:44 [ruaok]
I know we have some bugs that just get pushed ahead of ourselves.
19:47:49 [ruaok]
* ruaok nods at ocharles
19:48:07 [ruaok]
for bugs that simply seem to not get done, mention them in the meeting.
19:48:09 [warp]
ocharles: so they should really be set to "discussion required" ?
19:48:14 [ruaok]
put it on the agenda and we can talk about it.
19:48:21 [ocharles]
warp: or pushed into the agenda
19:48:21 [ocharles]
yea
19:48:31 [ijabz]
Yeah I can't think of an example at the moment, but I'll find one
19:48:48 [ocharles]
there are a few in the current milestone
19:49:01 [ruaok]
lets make an effort to pick at least one per meeting then.
19:49:14 [ocharles]
http://tickets.musicbrainz.org/secure/IssueNavigator.jspa?reset=true&jqlQuery=project+%3D+MBS+AND+fixVersion+%3D+%22Bug+fixes%2C+2012-07-09%22+AND+status+%3D+Reopened+ORDER+BY+priority+DESC three have been pushed into this release
19:49:42 [ruaok]
ok, then lets close the meeting for today.
19:49:49 [ruaok]
bug triage to follow as attention allows.
19:49:54 [ruaok]
thanks for your time!
19:49:55 [ocharles]
kewl
19:49:57 [ruaok]
</BANG>
19:49:57 [MBChatLogger]
MBChatLogger has changed the topic to: http://musicbrainz.org/#devel
19:50:07 [ruaok]
reosarevok: can you lead us into triage?
19:50:11 [warp]
thank you ruaok
19:50:12 [reosarevok]
5 min please
19:50:13 [reosarevok]
but yes
19:50:33 [ruaok]
k
19:50:34 [ocharles]
http://scheduling.ocharles.org.uk/results for the latest results
19:50:37 [ianmcorvidae]
ocharles: in those 5min, could you tell me the status of the varnish/caching stuff?
19:51:02 [ocharles]
ianmcorvidae: sure
19:51:29 [ianmcorvidae]
specifically, what's missing, code-wise and sysadmin-wise, before we can turn this on :)
19:51:34 [ocharles]
well, loads is missing
19:51:54 [ocharles]
the haskell daemon is capable of responding to changes to artist and recording rows, by simply adding bans to Varnish
19:52:10 [ocharles]
there is no work done on the musicbrainz-server side to correctly add those Dependencies: headers
19:52:24 [ocharles]
I hacked one in just to test, but not sensibly enough to even bother comiting
19:52:28 [ruaok]
bitmap: http://bitmap.mbsandbox.org/release/cac64a87-42f9-4c1c-a5ef-1e6824e20678/edit-relationships crashes if you're not logged in
19:52:59 [warp]
ocharles: wait, I don't get to use php but you do get to use haskell?
19:53:02 [ocharles]
ianmcorvidae: something I considered was changing all of the data access methods to log to $c all of the resources they have touched, and then dumping that into a Dependencies header
19:53:05 [ocharles]
warp: it's a prototype
19:53:08 [ocharles]
warp: it's not final
19:53:10 [warp]
:D
19:53:12 [ruaok]
lol
19:53:15 [ocharles]
as was yours ;)
19:53:44 [warp]
:P
19:53:51 [ianmcorvidae]
okay, that seems sane
19:54:00 [ocharles]
ianmcorvidae: README.md will hopefully have enough for you to get up and running and have a play
19:54:18 [ianmcorvidae]
yeah; I'll have to see about getting myself a postgres9.2 install but other than that :)
19:54:39 [ocharles]
yea, with pg 9+ there is pg_upgrade to easily upgrade to 9.2
19:54:47 [ocharles]
but i just did a clean reimport because my data was stale
19:54:48 [ruaok]
what requires 9.2?
19:55:00 [ianmcorvidae]
ruaok: row_to_json
19:55:10 [ianmcorvidae]
for sending a row to the invalidator daemon
19:55:17 [ocharles]
yea
19:55:19 [reosarevok]
* reosarevok returns, let me know when you're ready :)
19:55:20 [ianmcorvidae]
noting again, this is a prototype :)
19:55:24 [ruaok]
ok.
19:55:38 [ruaok]
I was just going to raise an objection to that.
19:55:45 [ocharles]
it would only be 9.2 on our servers
19:55:50 [ocharles]
it doesn't change dependencies anywhere else
19:56:26 [ianmcorvidae]
well, 9.2 is still very beta -- but maybe by the time this is done it won't be :P
19:56:33 [ocharles]
due 3rd quarter this year
19:56:42 [ianmcorvidae]
or someone will have backported an extension so older version can have row_to_json
19:57:48 [ocharles]
ruaok, reosarevok : triage then?
19:57:50 [reosarevok]
Sure
19:58:05 [ruaok]
yep
19:58:15 [reosarevok]
70% or more for unscheduled
19:58:15 [reosarevok]
IssueWithin 3 MonthsWithin 12 MonthsUnscheduled
19:58:15 [reosarevok]
[MBS-3871] Flag URLs with inappropriate content119
19:58:18 [reosarevok]
(only one there)
19:58:21 [ocharles]
fine
19:58:25 [reosarevok]
yeah
19:58:26 [ruaok]
plz
19:58:51 [reosarevok]
50% or more of the votes are for 12 months and 70% or more for either 3 months or 12 months:
19:58:55 [ianmcorvidae]
unscheduled or wontfix, possibly schema change :P
19:59:07 [reosarevok]
[MBS-3487] batch edit recordings from a release page372
19:59:29 [reosarevok]
This is useful, but it seems doable with scripts and easy to misuse by new editors :/
19:59:39 [reosarevok]
(I definitely want to use it though)
19:59:55 [ruaok]
12
19:59:56 [ocharles]
hmm
19:59:56 [reosarevok]
(either in-site or script-based)
20:00:06 [ocharles]
isn't this the release relationship editor?
20:00:12 [ocharles]
the ticket is pretty vague as to what it actually wants
20:00:32 [reosarevok]
ocharles: the ticket wants an easy way to follow our live recording guidelines
20:00:42 [reosarevok]
(which call for a specific string in the comment)
20:00:43 [warp]
oh, triage.
20:01:05 [reosarevok]
For that, there's a bitmap script already
20:01:12 [reosarevok]
I also want it to work for recording artists so I can follow our new CSG
20:01:20 [reosarevok]
For that there's nothing right now
20:01:22 [ocharles]
i'm ok with 12
20:01:37 [reosarevok]
If we want to do it, I say 12
20:01:54 [warp]
based on votes, 12.
20:02:17 [nikkimini]
nikkimini has joined #musicbrainz-devel
20:02:27 [reosarevok]
hi nikkimini
20:02:28 [ruaok]
hi nikkimini
20:02:30 [warp]
(but it sounds like it's ill-defined, so going to be a discussion required ticket as soon a developer looks at it properly)
20:02:32 [reosarevok]
Welcome to triage-time
20:02:37 [nikkimini]
hi. I can't tell the time :(
20:02:45 [reosarevok]
[MBS-3663] A negative vote should extend the expiration date of an edit371
20:02:53 [warp]
12
20:03:04 [ocharles]
hello more open edits!
20:03:05 [reosarevok]
Again: if we want this, 12
20:03:05 [ruaok]
post-edit system re-write.
20:03:10 [reosarevok]
(and I'm again not sure if we do)
20:03:16 [ocharles]
ruaok: so unscheduled?
20:03:19 [ruaok]
yeah
20:03:22 [ocharles]
+1
20:03:24 [warp]
yeah, that makes sense.
20:03:27 [ianmcorvidae]
+1 unsched
20:03:34 [warp]
it'll probably be invalid with NES
20:03:46 [reosarevok]
ocharles: well, the point is only to give it a bit of time for an editor to reply to a last-minute no vote
20:03:57 [reosarevok]
So it wouldn't cause a big change in the number of open edits, really...
20:04:01 [warp]
(yes, I'm going to consistently call the new edit system, NES)
20:04:31 [Prophet5]
Prophet5 has joined #musicbrainz-devel
20:04:46 [ocharles]
unsched it is
20:04:54 [reosarevok]
I can see why it is a pain, but it could always be changed to "make sure this doesn't happen in NES" :p
20:05:22 [reosarevok]
Anyway
20:05:23 [reosarevok]
Onwards
20:05:27 [reosarevok]
Standard list now:
20:05:28 [reosarevok]
[MBS-3072] Add JSON output to ws/2235
20:05:31 [ruaok]
3 months
20:05:38 [ruaok]
its part of NES
20:05:42 [ocharles]
warp: do you agree you can do this in 3 months?
20:05:51 [ocharles]
seeing as you are primarily working on that
20:06:01 [warp]
ocharles: sure, I can make time for it. I don't have anything else NES related on my plate yet.
20:06:12 [ocharles]
alright, 3 then
20:07:02 [warp]
* warp clicks assign to me.
20:07:24 [reosarevok]
[MBS-2138] Point people to the docs when loading /ws/2/ itself051
20:07:35 [reosarevok]
* reosarevok has no opinion since he's not a ws user
20:07:37 [ruaok]
thats trivial.
20:07:42 [djce]
djce has joined #musicbrainz-devel
20:07:46 [ocharles]
i'm fine with 12mo
20:07:49 [ruaok]
I'd say 3 months and get it out of the way
20:08:05 [ruaok]
but I dont feel strongly about it.
20:08:17 [ocharles]
if it's in 12, chances are I'll probably do it in the next few weeks when I'm bored
20:08:29 [ruaok]
fine.
20:08:31 [reosarevok]
And if it's in 3, chances are the same .p
20:08:33 [reosarevok]
*:p
20:08:36 [ocharles]
:)
20:08:51 [ruaok]
next
20:08:52 [reosarevok]
So I'd say 3 and if you don't do it in the next few weeks you can do it in 3 mo :p
20:08:57 [reosarevok]
but anyway
20:09:03 [reosarevok]
[MBS-3883] Add "undo edit" option145
20:09:05 [reosarevok]
NES
20:09:08 [reosarevok]
12?
20:09:10 [reosarevok]
No idea
20:09:10 [ruaok]
wontfix, NES
20:09:17 [reosarevok]
how is it a wontfix?
20:09:21 [ianmcorvidae]
do we have a NES ticket?
20:09:22 [reosarevok]
It is a fix, as part of NES :p
20:09:30 [ocharles]
a NES ticket would be nice
20:09:31 [ruaok]
because this is essentially part of NES.
20:09:32 [ianmcorvidae]
I'd say wontfix it and link it as duplicate
20:09:36 [ocharles]
and we can just keep saying "will be resolved by"
20:09:42 [ruaok]
ocharles: +1
20:09:48 [ocharles]
ok, i'll make that happen
20:09:51 [ruaok]
k
20:09:52 [warp]
* warp nods.
20:10:04 [reosarevok]
[MBS-4094] Make artist credits more compact in edits322
20:10:15 [ruaok]
fine
20:10:18 [reosarevok]
decision required I'd say...
20:10:34 [nikkimini]
who needs to make the decision?
20:10:36 [reosarevok]
As nikkimini commented "Artists credits are shown separately on purpose because it's not easy to distinguish artists on release pages"
20:10:50 [reosarevok]
Well, he needs to convince people that it was a bad decision :p
20:11:05 [nikkimini]
I also said "I do agree that it could be improved though" :P
20:11:22 [reosarevok]
Then the decision is how to improve it
20:11:32 [ocharles]
I'm ok with 3 months, but I need clear specs of what needs to be changed
20:11:36 [reosarevok]
(his option doesn't really work)
20:11:40 [ocharles]
designing improvements is outside my abilities, here
20:11:45 [nikkimini]
* nikkimini already said what she'd do in the comment
20:11:53 [ocharles]
and i'm fine with that, if that's all that needs to be done
20:12:06 [reosarevok]
I have no real issue with how it's now but I have no objection to nikkimini's idea either
20:12:44 [warp]
I have many issues with how it is now :)
20:13:13 [ruaok]
onward?
20:13:26 [reosarevok]
warp: please discuss them in the comment + suggest improvements then
20:13:33 [reosarevok]
* reosarevok watches ticket
20:13:34 [reosarevok]
[MBS-2491] Make it easier to add VA releases with artist aliases144
20:13:55 [ocharles]
where's 4094 going?
20:13:56 [ocharles]
3 months?
20:14:00 [warp]
reosarevok: I don't have any suggestions on how to improve it :)
20:14:14 [ocharles]
i'm tempted to say 12 just because this needs some talk
20:14:20 [warp]
12
20:14:22 [ruaok]
12
20:14:24 [reosarevok]
I'd agree with 12 if it's about 4094
20:14:31 [ocharles]
yea, i was still on 4094
20:14:38 [ruaok]
is 2491 impacted by NES?
20:14:48 [nikkimini]
no, it's interface stuff
20:14:51 [reosarevok]
ruaok: doubt it - it's release editor parsing
20:14:54 [ruaok]
k
20:15:13 [ruaok]
12 then
20:15:14 [warp]
I'm not sure I agree with the proposed change
20:15:32 [reosarevok]
warp: why? It'd make sense to default to the indicated credits
20:15:33 [warp]
(I'd say this needs some more discussion from UI people).
20:15:37 [ocharles]
it's a tricky one, indeed
20:15:45 [ocharles]
reosarevok: because most of the time I search for things in lower case
20:15:53 [warp]
reosarevok: most of the time I enter an artist merely to search for it
20:15:58 [nikkimini]
yeah, you would have to enter the exact name all the time
20:16:00 [ocharles]
and i generally dislike having specialist code for specific types of text
20:16:19 [reosarevok]
* reosarevok shrugs
20:16:21 [ocharles]
unsched + discussion
20:16:24 [reosarevok]
Not a change I care about too much
20:16:25 [warp]
reosarevok: I understand the problem, but I disagree with the solution.
20:16:33 [warp]
ocharles: +1
20:16:36 [reosarevok]
It can be annoying, but...
20:16:36 [bitmap]
bitmap has joined #musicbrainz-devel
20:17:05 [reosarevok]
I'm ok with unsched + discussion, but please do discuss it :p
20:17:10 [nikkimini]
I guess unsched until there's some sort of idea how it can be fixed without annoying people who are just searching
20:17:17 [reosarevok]
(post a comment with your problem with the current proposal)
20:17:17 [ocharles]
* ocharles nods
20:17:22 [reosarevok]
Ok
20:17:26 [reosarevok]
Onwards then
20:17:26 [reosarevok]
[MBS-2404] Private mb dump files to be encrypted015
20:17:37 [reosarevok]
* reosarevok has no idea how important this is
20:17:43 [ruaok]
this is a far out edge case.
20:17:53 [ocharles]
it's inline with SSL and stuff though
20:18:10 [ruaok]
because one a semi-regular basis I ride to digital west and change backup disks that have the private data on them
20:18:14 [warp]
encrypted with what?
20:18:26 [ocharles]
warp: that's an implementation detail
20:18:32 [ruaok]
if someone where to mug me, the backup disks could fall into someone else's hands.
20:18:40 [ocharles]
the goal is presumably to stop someone picking up mbdump-private and having a dump of passwords
20:18:55 [ruaok]
thus for that case if the files were encrypted no one could get the email addresses.
20:19:03 [ruaok]
(and other private info)
20:19:04 [warp]
what else is in -private?
20:19:08 [ocharles]
email addresses
20:19:10 [ocharles]
tags
20:19:11 [ocharles]
ratings
20:19:23 [ruaok]
warp: everything that isn't in the public dumps
20:19:28 [warp]
ok
20:19:41 [nikkimini]
* nikkimini would say 12 months
20:19:45 [ruaok]
[whole DB] - [public dumps] = [private dumps]
20:19:48 [ruaok]
ok with 12
20:19:51 [ocharles]
i mean this is probably a case of just piping the tar through some-encryption-program
20:19:53 [warp]
ok
20:19:56 [ocharles]
and doing the reverse on import
20:19:57 [ocharles]
so ok with 12
20:19:58 [ruaok]
ocharles: yep
20:20:05 [reosarevok]
12 seems OK - if it's very easy and someone wants to do it earlier feel free
20:20:11 [ruaok]
k
20:20:17 [reosarevok]
[MBS-3721] Change order of selecting cover art when multiple sources are available451
20:20:34 [reosarevok]
nikkimini, still valid?
20:20:40 [warp]
ocharles: well, probably with public/private keys and such. so as to not have the decryption keys on our servers, etc..
20:20:45 [nikkimini]
yes, but now we have the caa it might be a wontfixc
20:20:54 [ocharles]
+1 wontfix
20:21:02 [ruaok]
fine with wontfix
20:21:03 [warp]
+1 wontfix
20:21:04 [ocharles]
and barcodes will be going away, eventually
20:21:08 [reosarevok]
THREE MONTHS
20:21:13 [reosarevok]
Fiiine, wontfix :p
20:21:43 [reosarevok]
[MBS-3518] Remove chrome frame warning for IE9/10?512
20:21:51 [ocharles]
fine with 3
20:21:51 [ruaok]
seems clear. 3 mo
20:21:58 [nikkimini]
3
20:22:00 [ocharles]
though warp is the man with the vm's, so he might be the one to do that
20:22:17 [reosarevok]
Can't we remove it for all, and if they can't use MB it's their fault for being on IE8-? :p
20:22:22 [reosarevok]
But yeah
20:22:24 [warp]
3.
20:22:48 [warp]
reosarevok: this is related to my job of figuring out browser requirements and proposing something next week.
20:23:04 [reosarevok]
Cool
20:23:05 [reosarevok]
Onwards
20:23:11 [reosarevok]
[MBS-2431] Inherited ratings from works117
20:23:15 [nikkimini]
unsched
20:23:26 [warp]
+1
20:23:29 [reosarevok]
unsched with tendency to wontfix :p
20:23:29 [ruaok]
+1
20:23:33 [ianmcorvidae]
unsched +1
20:23:39 [ocharles]
fine
20:23:55 [reosarevok]
[MBS-750] Don't display "Medium 1" (or "CD 1", "Vinyl 1", ...) when there's only one medium in the release335
20:24:00 [reosarevok]
I'd say wontfix
20:24:01 [nikkimini]
wontfix
20:24:03 [ocharles]
wontfix
20:24:07 [ruaok]
ok
20:24:09 [nikkimini]
* nikkimini looks there for the format info and would get confused :P
20:24:15 [ruaok]
(with wontfix)
20:24:18 [ocharles]
and i like the consistency
20:24:21 [reosarevok]
Although if it gets us to remove the tracklist pages from public view, I'd accept it :p
20:24:31 [reosarevok]
* reosarevok hates those
20:24:40 [nikkimini]
I'd +1 removing the links to those :P
20:24:40 [reosarevok]
But yeah, probably wf
20:24:40 [warp]
+1 wontfix
20:24:48 [nikkimini]
but the actual text I'd like to stay where it is
20:24:50 [reosarevok]
Yeah
20:25:04 [reosarevok]
nikkimini: is there a ticket to stop showing those?
20:25:07 [reosarevok]
[MBS-2520] Warn and try to prevent DiscIDs from being orphaned when they're removed.335
20:25:09 [nikkimini]
not that I know of
20:25:18 [reosarevok]
I'd so wontfix this
20:25:26 [reosarevok]
But then, I'd kinda drop discIDs :p
20:25:36 [reosarevok]
No idea what others think
20:25:46 [nikkimini]
unsched for me
20:25:47 [ruaok]
unshed.
20:25:53 [ocharles]
unsched
20:25:59 [ianmcorvidae]
unsched works, I'd say it's worth fixing but not on any timescale
20:26:09 [ruaok]
personally, I think we should remove discids and just use fuzzy matching.
20:26:10 [warp]
* warp sides with reosarevok
20:26:29 [ocharles]
ruaok: +1
20:26:38 [ocharles]
but that's another can of worms
20:26:44 [ianmcorvidae]
that's a longer-term change, yeah
20:26:45 [reosarevok]
ianmcorvidae: that ticket seems to recommend wild guessing
20:26:45 [warp]
* warp nods.
20:26:45 [ruaok]
not for now.
20:26:49 [nikkimini]
big can of worms, lets move on :P
20:26:50 [ruaok]
maybe discuss that at the summit
20:26:52 [reosarevok]
I don't think it's worth fixing
20:26:56 [ianmcorvidae]
for now unsched this and we can talk about that later :)
20:26:59 [ocharles]
lets leave it open in unsched
20:26:59 [ocharles]
yea
20:27:01 [ruaok]
+1
20:27:32 [reosarevok]
ianmcorvidae: still means someone might decide to work on it, and I fear that chance :p
20:27:32 [reosarevok]
[MBS-3276] Show weekday for release dates255
20:27:39 [reosarevok]
screams "userscript" to me
20:27:45 [ruaok]
wontfix
20:27:52 [ianmcorvidae]
unsched or wontfix, userscript it
20:27:56 [ianmcorvidae]
agreed
20:27:57 [ocharles]
hm
20:28:03 [warp]
12
20:28:14 [nikkimini]
it would be nice if we marked dates in the html nicely so that user scripts can take advantage of it
20:28:14 [nikkimini]
(
20:28:14 [reosarevok]
There is a userscript already actually
20:28:20 [nikkimini]
(maybe we do, we don't for countries)
20:28:21 [reosarevok]
Or two
20:28:27 [ocharles]
I don't see why not, so I say unscheduled
20:28:32 [ocharles]
(and no wontfix)
20:28:35 [warp]
nikkimini: we should probably have some fancy rdfa for dates.
20:28:45 [ocharles]
microformats \o/
20:29:00 [reosarevok]
enough mental format masturbation
20:29:04 [ianmcorvidae]
:P
20:29:05 [reosarevok]
[MBS-941] Add RSS feeds157
20:29:10 [ruaok]
hack day.
20:29:13 [ruaok]
12 months
20:29:26 [ocharles]
we do need it soon, so i say 12 too
20:30:10 [warp]
that needs more discussion
20:30:10 [reosarevok]
nikkimini: there is now http://tickets.musicbrainz.org/browse/MBS-4931
20:30:12 [warp]
(but ok with 12)
20:30:18 [reosarevok]
on this ticket, 12 works
20:30:24 [ruaok]
* ruaok prods ocharles to post about the hack day
20:30:34 [ocharles]
ruaok: have you filled in the blanks?
20:30:41 [ianmcorvidae]
the hack day proposal is only part of that ticket
20:30:45 [ianmcorvidae]
but I agree 12m
20:30:59 [ruaok]
phone, sorry
20:31:04 [reosarevok]
[MBS-3326] Add the 'Tracks with too many capital letters' report046
20:31:07 [nikkimini]
wontfix
20:31:20 [ocharles]
unsched
20:31:24 [reosarevok]
either wontfix, or unsched but ignoring anything set to country:JP
20:31:50 [nikkimini]
there's no obvious use case for it really. the pre-ngs one was just a big mess
20:31:55 [warp]
wontfix
20:32:19 [ianmcorvidae]
what reo said, as far as I'm concerned; I certainly wouldn't use it so wontfix is fine by me
20:32:24 [ocharles]
ok, wontfix
20:33:13 [reosarevok]
[MBS-3328] Add the 'PUID collisions' report215
20:33:18 [reosarevok]
wont with a side dish of fix
20:33:19 [nikkimini]
nuke with fire
20:33:39 [ocharles]
haha
20:33:48 [ianmcorvidae]
wontfix, we're deprecating PUIDs soon anyway, aren't we? :P
20:33:48 [ocharles]
Fine with wontfix
20:33:52 [ocharles]
hopefully!
20:33:55 [warp]
puid? what's that?
20:34:15 [nikkimini]
it's broken, that's what :P
20:34:17 [reosarevok]
[MBS-3339] Always show artist disambiguation column next to it when editing a release751
20:34:33 [reosarevok]
Comment: This was fixed with http://codereview.musicbrainz.org/r/1669/ and MBS-2890 no?
20:34:38 [reosarevok]
( http://tickets.musicbrainz.org/browse/MBS-2890 )
20:34:40 [reosarevok]
Not sure if true
20:34:41 [ianmcorvidae]
* ianmcorvidae notes that we already have rdfa for dates, by the by
20:35:16 [ocharles]
i think my attention has about gone now
20:35:21 [ocharles]
if you want to carry on, feel free
20:35:25 [ocharles]
but 9 hours of work is about my limit :)
20:35:41 [ruaok]
* ruaok returns
20:35:48 [reosarevok]
[MBS-1376] Try to avoid non-Latin sortnames057
20:35:56 [ruaok]
but I'm supah hungry, so I'd prefer to carry on tomorrow.
20:36:03 [reosarevok]
This was mine, but I'm tempted to wontfix it and deprecate sortnames and names and just use aliases :p
20:36:03 [warp]
ianmcorvidae: I expected as much. I did notice some other oddities in the RDFa which eventually will have to be addressed.
20:36:15 [ocharles]
votes.org pushed to my server, i'll shuffle things around tomorrow
20:36:56 [warp]
alright, so we stop now, right?
20:37:09 [ruaok]
lets.
20:37:19 [reosarevok]
fine
20:37:20 [ruaok]
lets carry on tomorrow, standard meeting time.
20:37:32 [warp]
ok!
20:37:40 [nikkimini]
if I don't turn up, please ping me
20:38:01 [ruaok]
k
20:38:11 [ruaok]
* ruaok runs off for fudz
20:39:00 [nikkimini]
nikkimini has left #musicbrainz-devel
20:59:53 [djce]
djce has joined #musicbrainz-devel
21:28:03 [ijabz]
ijabz has joined #musicbrainz-devel
21:37:00 [ianmcorvidae]
warp, ocharles: you might be in bed, but when running tests I get: Can't locate object method "javascript" via package "HTML::FormHandler::Field::Text::1" at t/lib/t/MusicBrainz/Server/Form.pm line 50, <FILE> line 285.
21:37:14 [ianmcorvidae]
warp, ocharles: thoughts? wrong HTML::FormHandler version or something more sinister?
21:38:42 [kurtjx_]
kurtjx_ has joined #musicbrainz-devel
21:41:38 [plaintext]
ruaok: can you ping me when you have some time?
21:49:15 [ruaok]
plaintext: ping
21:49:30 [plaintext]
ruaok: pong
21:49:31 [plaintext]
hi
21:49:46 [plaintext]
So I have a problem connecting to the splunk server
21:49:57 [plaintext]
I'm not sure which port to connect to
21:50:10 [plaintext]
I even tried port scanning, but I didn't find anything
21:50:30 [ruaok]
only the web interface is exposed right now
21:51:05 [ruaok]
djce: ping
21:51:06 [plaintext]
Oh I see
21:51:13 [djce]
ruaok: pong
21:51:17 [ruaok]
hiya, djce
21:51:18 [plaintext]
So once my code is actually ran on the server it will work,
21:51:24 [ruaok]
plaintext: it should
21:51:25 [plaintext]
but I can't test on my VM now
21:51:33 [ruaok]
but it would be nice to see if we could get you access now.
21:52:03 [ruaok]
plaintext: do you know what port splunk runs on that you need to connect to?
21:52:17 [plaintext]
Default is 8089
21:52:20 [ruaok]
djce: lets chat about getting plaintext some more access to our splunk server.
21:52:28 [djce]
sure
21:52:57 [ruaok]
if we give plaintext vpn access and an account on pino, then he could access the splunk server, right?
21:53:17 [ruaok]
i'm ok with giving plaintext access to the VPN. I trust him enough to not shank us. :)
21:53:18 [djce]
as well as everything else, yes
21:53:35 [djce]
what access does he need that he doesn't already have?
21:53:50 [ruaok]
connecting to the splunk server on port 8089
21:53:54 [plaintext]
I only need to remotely acces the splunk server via the python SDK
21:54:18 [ruaok]
alternatively, could we allow access from rika to port 8089?
21:54:32 [ruaok]
plaintext could run his scripts on rika -- that might be an acceptable solution
21:54:50 [plaintext]
but wait, then rika's db would need to be changed
21:55:08 [plaintext]
...or not
21:55:15 [ianmcorvidae]
you can run it against the replicated DB, right?
21:55:16 [plaintext]
if I'm somehow accessing my own VM from there
21:56:02 [ruaok]
ianmcorvidae: he *could* use a replicated DB. as long as he doesn't change any existing tables.
21:56:13 [ianmcorvidae]
yeah
21:56:15 [ruaok]
but that is not so clean of a solution.
21:56:26 [ruaok]
a new instance of a DB might be preferrable.
21:56:30 [ianmcorvidae]
it's also totally possible to create a new DB, yeah
21:56:46 [ruaok]
but really, this all comes down to how to access splunk on the other side of the firewall
21:57:03 [ianmcorvidae]
I don't think his user currently has postgres access on rika -- ping me if I should set that up :)
21:57:11 [ruaok]
ianmcorvidae: k
21:59:26 [plaintext]
so what should I do now?
21:59:54 [ruaok]
wait until djce can sort out the proper method to give you access.
22:00:11 [plaintext]
Oh okay
22:00:20 [plaintext]
Can I bring up my other problem in the meantime?
22:00:24 [djce]
What did we settle on? connect to splunk from rika?
22:01:54 [ruaok]
djce: I would love to hear which solution you prefer
22:01:57 [ruaok]
plaintext: go
22:02:08 [plaintext]
So I pushed the new code to:
22:02:32 [plaintext]
https://github.com/metabrainz/musicbrainz-server-log-analysis/tree/d8ecd7b053f39be61521466f5e959927b5924770
22:02:45 [plaintext]
I made the diff, and tried to upload it to codereview
22:03:06 [plaintext]
I got a strange error: "The file "log_sanitizer.py" (revision ff118454582bbe54587eea65e6ac284889e47981) was not found in the repository"
22:03:26 [plaintext]
Then alastairp suggested I might have done the diff backwards, but I checked and it was made correctly
22:03:55 [plaintext]
I'm a bit lost
22:05:15 [ruaok]
can you post the command you used to make and the output of the diff to a pastebin?
22:05:28 [plaintext]
well it's a single line so I can paste it here
22:05:47 [Prophet5_]
Prophet5_ has joined #musicbrainz-devel
22:06:15 [plaintext]
git diff --full-index 4d96564 d8ecd7b > log_sanitizer_diff_4
22:07:07 [plaintext]
Oh the output too
22:07:12 [ruaok]
gimme a sec.
22:07:53 [plaintext]
http://pastebin.com/dfM6CaVv
22:08:49 [ruaok]
that looks ok.
22:09:21 [ruaok]
ianmcorvidae: any idea what might be going wrong on that submission to CR?
22:09:44 [ianmcorvidae]
I don't, not really
22:10:02 [ianmcorvidae]
previous diffs were made otherwise the same, just with a different second sha1?
22:10:07 [ruaok]
djce: if you have no opinions on this, can you please allow access to port 8089 from rika? that would be helpful.
22:13:05 [djce]
ruaok: sure. sorry, busy. will do.
22:13:06 [ijabz]
ijabz has joined #musicbrainz-devel
22:13:15 [ruaok]
np. thanks.
22:13:23 [djce]
also busy trying not to let wiley's disk fill up :-)
22:13:34 [ruaok]
yeah, what's that about?
22:13:46 [ruaok]
I went in and cleaned up a couple of ovas.
22:13:51 [djce]
web stats processing eats hugse disk space
22:13:54 [djce]
huge
22:14:05 [djce]
trying to fix that now...
22:14:11 [ianmcorvidae]
hm, revision 3 on codereview looks buggy, it looks like it's just the diff between 2 and 3 (rather than an actual third-revision diff)
22:14:33 [ruaok]
ianmcorvidae: maybe discard the review and start fresh?
22:14:35 [ianmcorvidae]
actually, likewise with the diff you just made
22:14:39 [ianmcorvidae]
now that I look at it
22:15:46 [ianmcorvidae]
discard + make anew seems reasonable
22:16:15 [ruaok]
plaintext: give that a try
22:16:16 [ianmcorvidae]
and make sure the diff includes everything to be reviewed every time -- the start commit for the diffs should generally always be the same
22:16:41 [plaintext]
I'm not sure I understand
22:17:00 [plaintext]
I mean
22:17:16 [plaintext]
so far I always made the diff between the new and the last versions
22:17:18 [plaintext]
is that wrong?
22:17:21 [ianmcorvidae]
yes
22:17:29 [plaintext]
I see
22:17:30 [ianmcorvidae]
you should always make it from the start of all the changes to the head
22:17:46 [ianmcorvidae]
so in your case it should always be dc3e5afb68..HEAD, basically
22:18:20 [plaintext]
dumb question: what exactly does "head" mean?
22:18:29 [ianmcorvidae]
what's currently checked-out
22:18:34 [ianmcorvidae]
whatever that happens to be
22:18:37 [plaintext]
thanks
22:19:02 [ianmcorvidae]
but yeah
22:19:12 [plaintext]
I'll discard the current review request then
22:19:12 [ianmcorvidae]
you're doing revisions of the diff, not things to append to the diff, is the way to think about it
22:19:35 [plaintext]
Oh I see
22:19:45 [ianmcorvidae]
i.e. the whole change should be "all of the current diff", not diff 1 + diff 2 + diff 3 ... etc.
22:19:58 [plaintext]
makes sense now
22:20:02 [ianmcorvidae]
cool :)
22:20:59 [ianmcorvidae]
ocharles: you need to update scheduling.ocharles.org.uk, I just got a fixed bug (http://tickets.musicbrainz.org/browse/MBS-4895) :)
22:25:35 [plaintext]
Okay so I did a diff between the initial version and the current and I get this:
22:25:37 [plaintext]
"The file 'README.md' (rc008427f71536793fb540997bddad2d78fb16280) could not be found in the repository"
22:25:49 [djce]
plaintext: firewall opened
22:25:52 [plaintext]
Is this because I didn't add the file in the commit?
22:25:55 [plaintext]
djce: thanks!
22:26:19 [plaintext]
Sorry for being such a newb with git
22:27:31 [ianmcorvidae]
plaintext: how'd you make the diff?
22:28:00 [plaintext]
git diff --full-index dc3e5afb68 d8ecd7b053 > new_diff
22:31:11 [ianmcorvidae]
hm
22:32:54 [ianmcorvidae]
I'm thoroughly confused at this point, then :/
22:34:11 [ianmcorvidae]
I'm wondering what codereview has for its repository
22:34:22 [plaintext]
all I did was: git init, git add log_sanitier.py, git commit ..., git push
22:34:34 [plaintext]
I picked musicbrainz-server-log-analysis
22:35:02 [ianmcorvidae]
yeah -- I mean, I wonder what codereview's mirror of that repo looks like
22:35:09 [ianmcorvidae]
if it failed to pull something or is pointing at the wrong repo or something
22:35:10 [Prophet5]
Prophet5 has joined #musicbrainz-devel
22:35:17 [plaintext]
I see
22:38:57 [ianmcorvidae]
hm
22:41:04 [ianmcorvidae]
the very first diff you put up, on the old review -- how did you make that?
22:42:01 [plaintext]
I don't really remember
22:42:39 [ianmcorvidae]
hm
22:49:20 [ianmcorvidae]
it does seem like something's misconfigured on reviewboard
22:49:23 [ianmcorvidae]
I'm not sure how to progress
22:50:15 [plaintext]
btw, thanks for helping me :)
22:50:25 [ianmcorvidae]
no problem, sorry I haven't been much help yet :P
22:51:39 [ianmcorvidae]
ruaok: any chance you can set me as an admin on reviewboard so I can look at it from that angle?
22:51:53 [ruaok]
sure
22:52:07 [ruaok]
incidentally, I'm working on answering the other question you asked me ianmcorvidae
22:52:15 [ianmcorvidae]
cool :)
22:53:21 [ruaok]
try now
22:53:26 [the_metalgamer]
the_metalgamer has joined #musicbrainz-devel
22:53:45 [ianmcorvidae]
I don't see anything different, hm
22:58:13 [ianmcorvidae]
ruaok: doesn't look like it took
22:58:32 [ruaok]
log out and back in?
22:58:48 [ianmcorvidae]
done it twice :)
22:59:57 [ianmcorvidae]
heh, looks like I can edit other people's codereviews though
23:00:11 [ianmcorvidae]
so I guess something changed, but not something that gives me access to look at repositories
23:01:41 [ruaok]
try now.
23:01:51 [ruaok]
previously i checked the "is superuser" checkbox.
23:02:07 [ruaok]
now I explicitly added all the perms to your account
23:02:14 [ianmcorvidae]
ah
23:02:27 [ianmcorvidae]
you should set me as "staff" according to http://www.reviewboard.org/docs/manual/1.5/admin/management/users/#staff-status
23:02:50 [ianmcorvidae]
superuser gives me the permissions but doesn't show the admin link, I guess :/
23:04:05 [ianmcorvidae]
which seems silly
23:04:06 [ianmcorvidae]
but anyway
23:04:16 [ruaok]
you have all teh statuses now.
23:05:46 [ianmcorvidae]
bah, different login at /admin/ that won't let me in; still no link on the main site
23:06:58 [ianmcorvidae]
we can leave it for now, you have other work to do
23:15:31 [warp]
ianmcorvidae: I have no idea about the FormHandler javascript stuff, haven't seen that before.
23:15:46 [ianmcorvidae]
warp: hm.
23:16:58 [kurtjx]
kurtjx has joined #musicbrainz-devel
23:17:02 [ianmcorvidae]
looks like it's coming from the $form->field('foo.baz.2.qux')->javascript ('alert("wait, wat?")'); line, I guess
23:17:35 [warp]
hm, in the tests.
23:17:45 [ianmcorvidae]
yeah, this is only when running tests
23:17:53 [warp]
meh, I also forgot how to check which version I have installed.
23:18:50 [ianmcorvidae]
perlmonks suggests perl -MMODULE -e 'print $MODULE::VERSION';
23:19:26 [hawke__]
*grumble* all these ISEs… :-(
23:19:26 [warp]
I don't get carton.
23:19:41 [warp]
carton exec -Ilib -- perl -MHTML::FormHandler -e 'print $HTML::FormHandler::VERSION'
23:19:44 [warp]
Can't locate HTML/FormHandler.pm in @INC
23:20:04 [warp]
so I'm obviously doing something wrong :)
23:20:25 [warp]
ianmcorvidae: just looking at the file, version is 0.36002
23:20:27 [ianmcorvidae]
heh, that looks right
23:20:27 [ianmcorvidae]
okay
23:20:31 [ianmcorvidae]
I have 0.40012
23:21:10 [warp]
so we may have to fix either our code or our Makefile.PL and carton.lock.
23:21:15 [ianmcorvidae]
yeah
23:21:20 [ianmcorvidae]
I'm actually not using carton, so that may be why
23:21:29 [warp]
aha!
23:21:30 [warp]
:)
23:21:38 [ianmcorvidae]
I just cpanm'd everything :P
23:21:53 [warp]
anyway. it's late here, so am off to bed now.
23:22:22 [warp]
goodnight #musicbrainz-developers
23:30:55 [reosarevok]
hawke__: I think the ISEs are a beta thing mostly so maybe you should temporarily move to normal mb
23:31:36 [ianmcorvidae]
* ianmcorvidae tries out tests again, installed an older version of HTML::FormHandler
23:42:48 [kyomi]
kyomi has joined #musicbrainz-devel
23:46:14 [kyomi]
I have some feedback I'd like to express on Picard 1.0 windows if this could be the place for it
23:51:13 [ianmcorvidae]
ocharles: looks like the origin/split-domains branch isn't pushed
23:51:29 [ianmcorvidae]
ocharles: I assume it's just my branch plus the merge though
23:57:07 [kurtjx]
kurtjx has joined #musicbrainz-devel