IRC log of musicbrainz-devel on 2013-10-17

Timestamps are in UTC.

00:03:52 [JesseW_not_logge]
I'm used to Gerrit, which prefers rebases.
00:04:51 [JesseW_not_logge]
btw, for https://bitbucket.org/metabrainz/musicbrainz-server/pull-request/500/mbs-1754-display-standalone-recordings-on -- now that I've gotten one Approve, what's the next step?
00:05:01 [mb-chat-logger]
http://tickets.musicbrainz.org/browse/MBS-1754
00:06:01 [ianmcorvidae]
well, so, it's sort of a matter of discretion on the part of the people who can merge it
00:06:24 [ianmcorvidae]
since ocharles looked at a previous diff I'm hoping to get an approve from him as well before merging it
00:07:06 [ianmcorvidae]
our cycle is one week of merge, one week of freeze, etc.
00:07:34 [JesseW_not_logge]
nods, that makes sense
00:07:35 [ianmcorvidae]
this week is a merge week, so ideally you get that approve then (or I decide it's fine and go ahead anyway)
00:08:00 [ianmcorvidae]
it then gets merged to beta and stays there until the next release, which is at the beginning of the next merge week (so, a week of freeze for testing on beta)
00:08:15 [ianmcorvidae]
so, yeah
00:08:42 [ianmcorvidae]
sometimes we'll merge things with one approve, sometimes we'll want more if someone was involved but hasn't approved yet, or there's a particular person we want to be sure is okay with the change
00:20:23 [JesseW_not_logge]
When adding a link to a translated string, should I do this way: l('Thank you, <a href="{editURL}">your edit</a> has been accepted and applied', { editURL => $c->uri_for_action('/edit/show', [ $edit->id ]) });
00:20:47 [ianmcorvidae]
ah, no, we have a syntax for links
00:20:57 [JesseW_not_logge]
excellent! what is it?
00:21:13 [JesseW_not_logge]
or, better, *where* is an example in the code?
00:21:45 [ianmcorvidae]
you want l('Thank you, {edit_url|your edit} has been accepted and applied', { edit_url => $c->uri_for_action(...etc...) })
00:21:57 [andreypopp]
andreypopp has joined #musicbrainz-devel
00:21:58 [ianmcorvidae]
see for example line 98 of root/layout.tt
00:22:16 [JesseW_not_logge]
thanks!
00:22:55 [JesseW_not_logge]
much cleaner, thanks.
00:23:00 [ianmcorvidae]
and remember that there can be more than one edit that note applies to, since I assume you're looking at the ticket asking that to link to the edit :)
00:24:04 [JesseW_not_logge]
yep; I'm implementing the single edit case first, as it's easier.
00:24:08 [ianmcorvidae]
(most edit pages are capable of producing several edits at once, and the hope is for more to do that as we get rid of having a bunch of edit pages for the same thing)
00:24:12 [ianmcorvidae]
fair enough!
00:24:42 [JesseW_not_logge]
But the next step is to figure out the edit search URL format, so I can link to a list of the other edits.
00:25:20 [JesseW_not_logge]
well, that's one next step -- the other one is to figure out how one edit is connected to the next, so I can do different things in Controller.pm if there are more than one edit...
00:25:49 [ianmcorvidae]
they largely aren't connected, as it happens, heh
00:26:50 [JesseW_not_logge]
yes, I thought that might be the case. If I have to, I can rely on the presence of an already existing flash message, and use that.
00:26:55 [JesseW_not_logge]
but that is very ugly, and I'd rather not. ;-)
00:31:49 [JesseW_not_logge]
ianmcorvidae: can you point me at a (hopefully simple-ish) edit page that can generate multiple edits?
00:35:43 [JesseW_not_logge]
hm. one of them is changing an artist name, and selecting the change artist credit checkbox, too.
00:36:04 [JesseW_not_logge]
but the 2nd edit in that case doesn't seem to be actually created through the usual method...
00:39:49 [JesseW_not_logge]
there seem to be only 15 calls to edit_action, but one of them is defining the edit method of the Edit Role, so the other uses probably come through there...
01:08:30 [MBJenkins]
Project full-import build #150: STILL FAILING in 7 min 28 sec: http://ci.musicbrainz.org/job/full-import/150/
01:18:02 [mrswag]
mrswag has joined #musicbrainz-devel
01:33:00 [ianmcorvidae|alt]
ianmcorvidae|alt has joined #musicbrainz-devel
01:42:25 [JesseW_not_logge]
JesseW_not_logge has joined #musicbrainz-devel
02:14:12 [misterswags]
misterswags has joined #musicbrainz-devel
02:40:09 [misterswags]
misterswags has joined #musicbrainz-devel
02:53:26 [CallerNo6]
CallerNo6 has joined #musicbrainz-devel
03:05:05 [CallerNo6]
from #musicbrainz:
03:05:07 [CallerNo6]
<tehsu-> I am seeing this error, [error] Caught exception in MusicBrainz::Server::Controller::Artist->load "Can't call method "name" on an undefined value at lib/MusicBrainz/Server/Data/Relationship.pm line 187." happens on some artists
03:44:05 [misterswag]
Use of uninitialized value in concatenation (.) or string at lib/MusicBrainz/Server/Entity/Artwork.pm line 80
03:46:50 [ruaok]
ruaok has joined #musicbrainz-devel
03:47:13 [misterswags]
misterswags has joined #musicbrainz-devel
04:03:38 [JesseW_not_logge]
JesseW_not_logge has joined #musicbrainz-devel
04:36:01 [dukduk_]
dukduk_ has joined #musicbrainz-devel
04:36:14 [dukduk_]
anyone on?
04:36:47 [Leftmost]
If you have a question, go ahead and ask it.
04:37:17 [dukduk_]
I've done some poking around on amazon and figured out some info about cover art...
04:37:54 [dukduk_]
I added a link to the "How to add cover art" Page on the wiki, but since it's my first edit, I wasnted to make sure i'm not stepping an any toes here.
04:37:57 [dukduk_]
http://wiki.musicbrainz.org/How_to_Add_Cover_Art
04:38:52 [dukduk_]
I stuck it in the further info section at the bottom since "how Tos" seem to be reserved for more user-friendly guides.
04:40:55 [dukduk_]
In any case, it's really more technical stuff (messing around with HTTP POST variables and whatnot that i thought devs would be more interested in than the average user.
04:40:55 [Leftmost]
Not really my area, but I'm sure that if there's an issue someone will correct it.
04:42:08 [voiceinsideyou]
voiceinsideyou has joined #musicbrainz-devel
04:43:24 [dukduk_]
...So i thought i'd bring it to you guys attention. I discovered a way to programatically pull fully zoomed (1448x1448) covers art from Amazon, but they're split into tiles. Some simple php img() funcs could rassemble them though and update all the cover art in the database to high-res if someone wanted to write a script.
04:43:55 [dukduk_]
for anyone interested, the new wiki page is @ http://wiki.musicbrainz.org/Amazon_Cover_Art
04:44:46 [voiceinsideyou1]
voiceinsideyou1 has joined #musicbrainz-devel
04:57:56 [voiceinsideyou]
voiceinsideyou has joined #musicbrainz-devel
05:05:30 [voiceinsideyou1]
voiceinsideyou1 has joined #musicbrainz-devel
05:35:54 [misterswag]
misterswag has joined #musicbrainz-devel
05:57:00 [mrswag]
mrswag has joined #musicbrainz-devel
06:04:14 [UmkaDK]
UmkaDK has joined #musicbrainz-devel
06:07:13 [voiceinsideyou]
voiceinsideyou has joined #musicbrainz-devel
06:11:30 [UmkaDK]
UmkaDK has joined #musicbrainz-devel
06:17:23 [rvedotrc]
rvedotrc has joined #musicbrainz-devel
07:44:35 [UmkaDK]
UmkaDK has joined #musicbrainz-devel
08:00:14 [marcooliveira]
marcooliveira has joined #musicbrainz-devel
08:03:04 [mrswag]
mrswag has joined #musicbrainz-devel
08:05:48 [ijabz]
ijabz has joined #musicbrainz-devel
08:11:08 [ijabz]
ijabz has joined #musicbrainz-devel
08:25:42 [marcooliveira]
marcooliveira has joined #musicbrainz-devel
08:26:52 [ijabz]
nikki transclude please http://wiki.musicbrainz.org/XML_Web_Service/Version_2/Search/
08:58:58 [reosarevok]
reosarevok has joined #musicbrainz-devel
09:09:59 [ijabz]
reosarevok standalone recording seach for just a documentation problem, btw you need to add places to http://wiki.musicbrainz.org/XML_Web_Service
09:11:49 [reosarevok]
ijabz: I saw the ticket comment, thanks! :)
09:12:05 [reosarevok]
And yeah, that sounds necessary, I'll do it in a while
09:12:16 [reosarevok]
(if I haven't done it by tomorrow, shout at me :) )
09:12:21 [UmkaDK]
Just a quick progress updated on Pg-9.0 issues
09:13:48 [UmkaDK]
… it seems that we've managed to get schema 19 running on Pg-9.0 by installing the database from the released dump files.
09:14:32 [UmkaDK]
… slave server seems to be running and replicating (it's too early to say for sure)
09:15:16 [UmkaDK]
… BUT it also seems that editor dumps are corrupted somehow.
09:17:25 [reosarevok]
Missing ha1?
09:17:36 [UmkaDK]
importing editor dumps fails at about 70% :(
09:17:40 [reosarevok]
Oh
09:17:48 [reosarevok]
So something else I guess
09:17:56 [reosarevok]
* reosarevok can't help with that, hopefully ocharles can :)
09:18:27 [UmkaDK]
no worries, we don't actually need those to run in slave mode, so not a big deal
09:18:50 [UmkaDK]
but I imagine it might crop up sooner or later! :)
09:19:12 [UmkaDK]
And thanks for all the help guys! Much appreciated!!
09:32:43 [Freso]
https://beta.musicbrainz.org/statistics : "Places: 0" :(
09:33:44 [andreypopp]
andreypopp has joined #musicbrainz-devel
09:42:24 [Freso]
ianmcorvidae: How does Geordi handle many-{many|one} Discogs-MB links?
09:58:24 [Hadora]
I get a lot of 404 with coverartarchive.org
09:58:35 [Hadora]
do you have information about that ?
09:59:28 [Hadora]
one example between many others : http://coverartarchive.org/release/12a7b341-1ab9-43e7-8180-67b941fc9fa2/
09:59:45 [Hadora]
while there is actually a cover http://musicbrainz.org/release/12a7b341-1ab9-43e7-8180-67b941fc9fa2/cover-art
10:01:10 [Hadora]
Should I use another method, like querying directly with the cover art id like this http://coverartarchive.org/release/12a7b341-1ab9-43e7-8180-67b941fc9fa2/5440692442.jpg ?
10:06:24 [kuno]
Hadora: hm, if you get a 302 from http://coverartarchive.org/release/12a7b341-1ab9-43e7-8180-67b941fc9fa2/ the file should be available.
10:06:55 [kuno]
Hadora: so that does like something is wrong on our end.
10:07:03 [Hadora]
I get a 404 actually
10:07:33 [Hadora]
and the message from caa is clear :
10:07:34 [Hadora]
Page not found
10:07:34 [Hadora]
We’re sorry, the page you have requested is not available.
10:07:35 [Hadora]
sry
10:07:59 [kuno]
Hadora: I mean, when you request http://ia601004.us.archive.org/28/items/mbid-12a7b341-1ab9-43e7-8180-67b941fc9fa2/index.json
10:08:03 [kuno]
woops
10:08:29 [kuno]
Hadora: when you request http://coverartarchive.org/release/12a7b341-1ab9-43e7-8180-67b941fc9fa2/ you get a 302 response, which is a redirect to the internet archive
10:09:08 [kuno]
Hadora: if you get a 404 at http://coverartarchive.org/release/12a7b341-1ab9-43e7-8180-67b941fc9fa2/, we do not have cover art. if you get a 302 redirect, we DO have cover art. in your case the URL we redirect to does end up at a 404, and that is a bug.
10:09:21 [ocharles]
Mornin'
10:09:50 [kuno]
ocharles: see Hadora's question, is our generating of index.json files queued up?
10:10:16 [ocharles]
Lessee
10:10:23 [Hadora]
it does work for some release http://coverartarchive.org/release/f6758afb-ecff-410e-8979-b131cfa3c6d3
10:10:28 [reodroid]
reodroid has joined #musicbrainz-devel
10:10:35 [Hadora]
so everything is ok on my side in that point of view
10:10:39 [ocharles]
2013-10-16 12:21:27.981255500 Wed Oct 16 12:21:27 UTC 2013 : Starting CAA-indexer
10:10:44 [ocharles]
with nothing else in the log after
10:10:49 [ocharles]
so that doesn't look too good
10:11:06 [kuno]
ocharles: are you looking into it? or would you like me to enter a ticket?
10:11:12 [ocharles]
i'm looking into it
10:11:20 [kuno]
great, then I go back to work :)
10:11:26 [ocharles]
there's a steady '0' for index events over the last 12 hours, meep
10:21:52 [ocharles]
only 3.5k messages to get through now
10:23:00 [marcooliveira]
marcooliveira has joined #musicbrainz-devel
10:23:58 [natethomas]
natethomas has joined #musicbrainz-devel
10:26:19 [ocharles]
Hadora: ok, it looks like things are back to operating correctly, but we have two days of backlog to catch up on
10:26:22 [ocharles]
this will take a while
10:26:41 [Hadora]
gotcha
10:27:22 [kuno]
ocharles: the two days is the period of missed events, not the time it takes to catch up, correct? :)
10:28:08 [ocharles]
2 days worth of events = 4k events. RabbitMQ reports we're processing 0.14 msgs/sec
10:28:30 [ocharles]
so ~1/3 days to process this
10:35:05 [Freso]
* Freso is about to add some to the queue
10:35:09 [MBJenkins]
Project caa-admin build #1: FAILURE in 2.2 sec: http://ci.musicbrainz.org/job/caa-admin/1/
10:35:23 [Freso]
ocharles: What the status on place/create?
10:35:35 [ocharles]
Freso: it's disabled? :)
10:36:15 [ocharles]
Freso: I still haven't fixed the seg fault, and until I do it will remain off
10:36:24 [MBJenkins]
Project caa-admin build #2: STILL FAILING in 2 sec: http://ci.musicbrainz.org/job/caa-admin/2/
10:36:24 [MBJenkins]
Oliver Charles: Relax base dependency to build on GHC 7.4
10:37:11 [Freso]
ocharles: I get that. I was more asking for a status report. :) You managed to recreate the segfault locally last night?
10:37:32 [MBJenkins]
Project caa-admin build #3: STILL FAILING in 1.8 sec: http://ci.musicbrainz.org/job/caa-admin/3/
10:37:44 [ocharles]
no
10:37:52 [ocharles]
I can still only cause the seg fault to happen on totoro
10:38:07 [ocharles]
but now that I have steps to reproduce it on totoro, I will try another clone of totoro and see if I can cause it again
10:38:13 [Freso]
right.
10:38:15 [Freso]
Weird.
10:39:39 [UmkaDK]
UmkaDK has joined #musicbrainz-devel
10:40:20 [MBJenkins]
Project caa-admin build #4: STILL FAILING in 0.8 sec: http://ci.musicbrainz.org/job/caa-admin/4/
10:40:57 [JonnyJD_]
Sebastinas: is the webbrowsers python module also patched for debian to include "iceweasel"? In that case you don't need to maintain that browser patch for isrcsubmit
10:42:09 [Hadora]
I'll go with http://coverartarchive.org/release/f6758afb-ecff-410e-8979-b131cfa3c6d3/front meanwhile
10:42:14 [Hadora]
which works
10:42:14 [JonnyJD_]
Sebastinas: also including python-discid as an alternative to python-libdiscid would be nice (packaged in the MusicBrainz PPA)
10:42:41 [Freso]
Hadora: What is it for?
10:43:02 [Hadora]
I think you get the first front cover
10:43:12 [MBJenkins]
Project caa-admin build #5: ABORTED in 27 sec: http://ci.musicbrainz.org/job/caa-admin/5/
10:43:21 [Freso]
Yeah. :) I meant, what are you using it for?
10:43:25 [JonnyJD_]
Sebastinas: PPA/backport for isrcsubmit is here: http://bazaar.launchpad.net/~jonnyjd/isrcsubmit/packaging/files
10:43:39 [nikki]
ijabz: I don't see any changes to transclude
10:44:00 [Hadora]
because CAA-indexer has some issues atm
10:44:43 [nikki]
front will give you the first thing marked as front, yes
10:44:50 [Hadora]
see above
10:45:08 [Freso]
Hadora: No, I mean. Are you using it for some project, or just for personal use? Why do you need to get the front cover?
10:48:40 [Hadora]
We use it for our website, why do you ask ?
10:50:02 [Hadora]
Do you mean there should be special rules for projects ?
10:50:51 [Freso]
No.
10:50:56 [Freso]
I'm just curious.
10:50:59 [Hadora]
oh ok
10:51:20 [Freso]
It's interesting to follow where MB is being used "in the wild".
10:51:24 [Hadora]
MB is our backbone for our music section at senscritique
10:51:26 [Hadora]
http://www.senscritique.com/musique
10:51:28 [Hadora]
french website
10:51:44 [Hadora]
and yes we do consider donating :D
10:51:52 [Freso]
:)
10:54:26 [zas]
zas has joined #musicbrainz-devel
10:54:33 [Freso]
Hadora: You import the images from CAA and then store (and edit) them locally and display from there?
10:54:40 [zas]
Hi
10:55:40 [Hadora]
we do, mainly not to stress other people's servers
10:56:02 [zas]
CAA didn't generate index.json for https://musicbrainz.org/release/273e233e-d0aa-4f60-b8f2-482b0c2e0a87 cover art, any idea why ?
10:56:10 [nikki]
there's a backlog
10:56:25 [prog_]
prog_ has joined #musicbrainz-devel
10:57:17 [MBJenkins]
Project caa-admin build #6: STILL FAILING in 12 min: http://ci.musicbrainz.org/job/caa-admin/6/
10:58:28 [ocharles]
boo
11:06:47 [Freso]
Wasn't there a WikiPage for collecting info on projects using MB in some way? I remember it as being a subpage of User:RobertKaye, but I can't find it in neither his nor my own contribution history, and it's not apparent from searching either.
11:06:55 [Freso]
Though, I guess perhaps it was deleted.
11:06:57 [Freso]
Ah well.
11:09:06 [reodroid]
Unless you mean MetaBrainz:Customers, I don't know it
11:11:19 [Freso]
reodroid: I don't mean customers, no, since e.g. Hadora's site isn't a customer (yet... :)). I think Rob made a collection of things using MB besides just customers. But maybe he deleted it again.
11:11:35 [Freso]
I'll try and remember to ask him when he's around.
11:12:56 [nikki]
wouldn't such a list end up being huge anyway?
11:13:35 [Freso]
Wouldn't that be great?
11:14:38 [nikki]
you could just link to a google search for "musicbrainz test artist" -site:musicbrainz.org and save us the trouble of maintaining it :P
11:17:34 [reodroid2]
reodroid2 has joined #musicbrainz-devel
11:18:35 [Freso]
nikki: That's 11 results with DuckDuckGo and 2 pages with Google. That's not that many. And for Google, some sites are listed more than ones - and it looks like some have MB Test Artist from a "3rd party source". (E.g., it looks like there are some downloads with that artist...)
11:18:46 [nikki]
2 pages?
11:18:50 [Freso]
nikki: And Hadora's site isn't listed.
11:19:22 [nikki]
I get far more than 2 pages
11:19:31 [Freso]
Oh. Google thought it should omit some results to only show the most relevant.
11:20:18 [nikki]
* nikki sighs at all the ISEs
11:20:19 [reodroid2]
reodroid2 has joined #musicbrainz-devel
11:20:37 [Freso]
This only created even more dupliecate results per site.
11:21:31 [misterswag]
misterswag has joined #musicbrainz-devel
11:21:40 [MBJenkins]
Project caa-admin build #7: ABORTED in 16 sec: http://ci.musicbrainz.org/job/caa-admin/7/
11:21:40 [MBJenkins]
Oliver Charles: Relax acid-state dependency to use bytestring >= 0.9.1.8
11:22:51 [Freso]
And I expect there are more false positives as well, and that some sites will also be missing (if they, e.g., manually block the test artist or only show data based on certain criteria (e.g., country, specific artists, ...).
11:24:07 [nikki]
anyway, I was joking there. that doesn't make my point any less valid, there are lots of sites and I really really doubt we have enough people keen enough on keeping it up to date and I also really doubt a huge long list of pages is going to be useful to anyone. we already have the customers page if we want something to impress people with
11:38:28 [Hadora]
We would like being listed somewhere as users of course
11:39:40 [andreypopp]
andreypopp has joined #musicbrainz-devel
11:45:59 [ocharles]
lunch!
12:01:14 [ijabz]
nikki, oh I thought it did all sections on the page, but its just this part that has changed http://wiki.musicbrainz.org/index.php?title=XML_Web_Service/Version_2/Search/RecordingSearch
12:06:17 [reosarevok]
reosarevok has joined #musicbrainz-devel
12:08:19 [kepstin-laptop]
kepstin-laptop has joined #musicbrainz-devel
12:52:08 [MBJenkins]
Project caa-admin build #8: ABORTED in 1 min 14 sec: http://ci.musicbrainz.org/job/caa-admin/8/
12:52:08 [MBJenkins]
Oliver Charles: Relax containers dependency
13:05:47 [night199uk]
night199uk has joined #musicbrainz-devel
13:09:52 [MBJenkins]
Yippie, build fixed!
13:09:53 [MBJenkins]
Project caa-admin build #9: FIXED in 15 min: http://ci.musicbrainz.org/job/caa-admin/9/
13:11:06 [ocharles]
at last
13:12:24 [MBJenkins]
Project caa-admin build #10: SUCCESS in 9.1 sec: http://ci.musicbrainz.org/job/caa-admin/10/
13:35:22 [ocharles]
oh ffs, I thought hobbes was 12.04
13:49:12 [reosarevok]
reosarevok has joined #musicbrainz-devel
13:58:00 [ocharles]
yessss, I am able to segfault a Pg database that is not production
13:58:45 [nikki]
:o
13:58:53 [nikki]
what did you do?
14:00:51 [ocharles]
just cloned totoro again and used my steps to reproduce it on prod
14:00:51 [ocharles]
so I might be able to get gdb hooked up and make some progress
14:01:01 [nikki]
* nikki wonders what the steps were :P
14:01:03 [ocharles]
I can segfault to my heart's content now
14:01:05 [nikki]
haha
14:01:17 [ocharles]
just insert into place, then update edits_pending
14:01:24 [ocharles]
the insert succeeds, the update doesn't
14:01:41 [nikki]
huh wow
14:04:40 [natethomas]
natethomas has joined #musicbrainz-devel
14:05:34 [reosarevok]
er. Weird.
14:09:01 [ocharles]
you say weird - it's a seg fault :p
14:09:37 [reosarevok]
I'm not even sure what that means :(
14:09:38 [nikki]
you can have weird seg faults and not-so-weird seg faults you know :P
14:09:44 [reosarevok]
* reosarevok feeling computer-dumb :)
14:09:54 [ocharles]
http://en.wikipedia.org/wiki/Segmentation_fault
14:13:06 [ocharles]
http://en.wikipedia.org/wiki/File:Card_reader_segfault.jpg also, lol
14:15:50 [reosarevok]
:)
14:21:47 [ocharles]
more progress - i'm now in gdb, looking at the source code of dbmirror itself
14:24:21 [misterswag_]
misterswag_ has joined #musicbrainz-devel
14:25:05 [misterswag_]
how come the server isnt accessible after i upgrade.sh and change lib/dbdefs.om and git checkout ;(
14:26:29 [misterswag_]
@40000000525f42ac180cfd74 [warning] GET http://192.168.1.4:5000/ caused a warning: Use of uninitialized value in concatenation (.) or string at lib/MusicBrainz/Server/Entity/Artwork.pm line 80.
14:35:32 [ocharles]
i has found the null!
14:35:38 [ocharles]
now to work out why it's null
14:36:16 [Freso]
ocharles: \o/
14:36:29 [reosarevok]
\o/
14:36:30 [Freso]
s/o/null/
14:36:45 [ocharles]
wait, maybe i haven't
14:36:49 [Freso]
Aw.
14:36:49 [ocharles]
i have found a null though :P
14:36:53 [Freso]
Heh.
14:37:18 [ocharles]
it does look like this is the offending line though
14:37:18 [ocharles]
iRetValue = SPI_execp(pplan, planData, NULL, 1);
14:37:33 [ocharles]
for the third parameter: If nulls is NULL then SPI_execp assumes that no parameters are null.
14:37:37 [ocharles]
yet planData[1] is NULL
14:37:59 [reosarevok]
* reosarevok is really happy someone here knows what he's talking about :)
14:38:19 [reosarevok]
(I certainly would have had absolutely no idea how to go around this :/ )
14:38:39 [reosarevok]
(even if I had found the issue on testing, I mean)
14:38:42 [ocharles]
y
14:38:49 [ocharles]
note to self:this window is not gdb
14:39:00 [Freso]
;)
14:40:06 [ocharles]
ok, want to know what the problem is
14:40:08 [ocharles]
?
14:40:14 [ocharles]
someone tell me what the primary key of place is
14:40:15 [reosarevok]
Ideally, yes
14:40:35 [reosarevok]
.....
14:40:40 [reosarevok]
* reosarevok hides under a table
14:40:54 [reosarevok]
* reosarevok goes check the code :p
14:41:04 [ijabz]
just looked up bmirror http://dbmirror.projects.pgfoundry.org/ sounds like it doesnt get much tlc these days
14:41:17 [ocharles]
ijabz: we have our own fork of it
14:41:23 [ocharles]
(which gets equally little tlc)
14:41:52 [ijabz]
doesnt like the kinfd of thing we want to be maintaining
14:42:33 [reosarevok]
Hmm
14:42:45 [reosarevok]
ocharles: " id SERIAL, -- PK
14:42:45 [reosarevok]
" is that not enough? :/
14:42:58 [ocharles]
reosarevok: there is no primary key in the upgrade scripts
14:44:30 [reosarevok]
So how do you indicate a PK on those?
14:44:46 [reosarevok]
* reosarevok obviously missed some obvious bit
14:45:18 [reosarevok]
https://bitbucket.org/metabrainz/musicbrainz-server/src/e4e202c1feba02d318d1f67f59692c2510a45194/admin/sql/updates/20130618-places.sql?at=beta does have that line which I thought did it, but I guess not then :/
14:45:37 [reosarevok]
It has to say "PRIMARY KEY" instead?
14:45:42 [ocharles]
either id SERIAL PRIMARY KEY
14:45:42 [ocharles]
-- is just a comment
14:45:42 [ocharles]
we parse the .sql file to generate CreatePrimaryKeys.sql
14:45:54 [ocharles]
yes, it has to be part of the SQL, not a comment
14:45:58 [ocharles]
or you can ALTER TABLE place ADD PRIMARY KEY (id)
14:45:59 [reosarevok]
:(
14:46:03 [reosarevok]
Ok
14:46:07 [reosarevok]
So how do we fix this?
14:46:17 [reosarevok]
Just run ALTER TABLE on it?
14:46:21 [ocharles]
i fix dbmirror to not take the entire database down, and I will add a primary key
14:46:51 [reosarevok]
Guess we'll need to tell other people to do that too?
14:46:57 [ocharles]
i added a primary key on my replica and it works
14:46:57 [ocharles]
musicbrainz_db_20110516=> UPDATE place SET edits_pending = numeric_larger(0, edits_pending + '1') WHERE id IN (828);
14:47:01 [ocharles]
ERROR: table lacks primary key
14:47:02 [ocharles]
to
14:47:06 [ocharles]
we are the only people who use dbmirror
14:47:07 [ocharles]
no*
14:47:08 [ocharles]
it's a master thing, not a slave thing
14:47:10 [reosarevok]
Oh
14:47:11 [reosarevok]
Ok
14:47:13 [ocharles]
oh, you mean creating the PK?
14:47:16 [reosarevok]
Yes :p
14:47:16 [ocharles]
that doesn't really matter
14:47:21 [reosarevok]
It doesn't?
14:47:23 [ocharles]
it's just an index, at the end of the day
14:47:28 [reosarevok]
* reosarevok shrugs
14:47:34 [reosarevok]
I'll trust you then
14:47:41 [reosarevok]
* reosarevok feels dumb now
14:47:54 [ocharles]
you're highlighting why i'm unhappy with these scripts though
14:48:01 [ocharles]
there's too much "I better remember to" shit
14:48:02 [mrswag]
mrswag has joined #musicbrainz-devel
14:48:07 [reosarevok]
(also sad about the code review, but obviously that wouldn't be an issue if I hadn't been dumb to begin with)
14:48:08 [ocharles]
it doesn't need to be this hard
14:48:50 [reosarevok]
Well, I guess that adds me to the club of "fuck up on schema changes" people
14:48:56 [mrswag]
What fires are left to be put out before next VM? Didn't know it was that diff to update?
14:49:01 [reosarevok]
Not the first one at least! :p
14:49:17 [ocharles]
reosarevok: this stuff is hard. as far as I'm concerned, we all fucked up
14:49:26 [nikki]
* nikki thinks we could say "it doesn't need to be this hard" about the interface too :P
14:49:27 [ocharles]
code review failed us, the tools failed us
14:49:54 [kuno]
reosarevok: welcome!
14:50:11 [reosarevok]
kuno: do I get a member card?
14:50:49 [kuno]
:)
14:51:02 [nikki]
reosarevok: so demanding! :P
14:51:27 [ocharles]
http://musicbrainz.org/place/4352063b-a833-421b-a420-e7fb295dece0
14:51:29 [ocharles]
after you, Freso
14:51:34 [ocharles]
or not ;)
14:51:57 [ocharles]
there's still a fairly critical bug that it doesn't work unless you enter an address
14:52:01 [ocharles]
but that's an ISE, not a seg fault
14:52:22 [mrswag]
The server?
14:52:32 [reosarevok]
Huh
14:52:33 [ocharles]
oh good, there's no PK on the relationship tables either :)
14:52:42 [reosarevok]
Yeah, I assume there's the same on all tables :(
14:53:20 [reosarevok]
But the address ISE is also something we didn't see in rika
14:53:21 [reosarevok]
What is it?
14:53:27 [Freso]
ocharles: I still get 503's on place/create
14:53:31 [Freso]
*503'd
14:53:38 [ocharles]
ok, one moment
14:53:47 [ocharles]
urgh, no foreign keys to place either
14:53:50 [Freso]
(@beta, of course)
14:54:32 [reosarevok]
I mean, everything that implies changing comments to non-comments will probably be borked
14:56:34 [ocharles]
Freso: ok, I haven't unblocked beta
14:56:45 [Freso]
mrswag: It's only really difficult to update the server code twice a year.
14:56:52 [Freso]
mrswag: Mid-May and Mid-October.
14:57:20 [Freso]
mrswag: Unfortunately for you, this update is the Mid-October one. :)
14:57:29 [mrswag]
Freso: yeah I thought I did everything right for the VM and server isn't accessible lol
14:59:56 [ocharles]
Freso: should work on beta now
15:00:26 [nikki]
hm. the addresses ISE is a bit annoying. that means I can't run my script to add a bunch of places I collected :P
15:00:45 [reosarevok]
What is it ISEing on?
15:00:58 [reosarevok]
* reosarevok doesn't want to go try and send nikki more mail :p
15:02:02 [nikki]
address being empty
15:03:10 [Freso]
nikki: But at least places *can* be entered now. As opposed to, say, an hour ago where it was entirely impossible. The ISE can (and will, I'm sure) be fixed too, and you can run your script then. :)
15:03:19 [reosarevok]
nikki: yes, I mean what file
15:03:44 [nikki]
Freso: but the longer I have to wait, the more likely it is to add dupes :P
15:04:01 [nikki]
I could just enter something as the address if it's blank... hmm
15:04:02 [reosarevok]
It doesn't have any Estonian places, right nikki? :p
15:04:08 [nikki]
no, because you didn't add any :P
15:04:30 [reosarevok]
Good
15:04:41 [reosarevok]
Then I can safely enter them without fear of dupes :)
15:04:43 [ianmcorvidae]
heh, it was a PK?
15:04:45 [ianmcorvidae]
crazy
15:04:55 [nikki]
https://musicbrainz.org/place/7a2a91c8-1c5b-4564-90fa-00290a6db009 :D
15:05:01 [ianmcorvidae]
I didn't even realize that the stuff dbmirror does with PKs actually worked :P
15:05:01 [ocharles]
ianmcorvidae: https://github.com/metabrainz/dbmirror/pull/1/files
15:06:18 [ianmcorvidae]
presumably there's a second fix to add the PK, as well, but that's presumably just done already on master?
15:06:27 [CallerNo6]
CallerNo6 has joined #musicbrainz-devel
15:06:37 [ocharles]
no, I just did that
15:06:44 [ocharles]
copy and pasted from Create*
15:06:52 [ianmcorvidae]
yeah, that's what I meant
15:08:57 [reosarevok]
"DBD::Pg::st execute failed: ERROR: unrecognized function API version: 6579570 at lib/Sql.pm line 107, line 3.
15:08:58 [reosarevok]
" le fuck
15:09:11 [ocharles]
ok, I think that's from me
15:09:14 [ocharles]
I have no idea how to fix that
15:09:24 [ocharles]
ok, yes I do
15:09:24 [ocharles]
one moment
15:09:28 [reosarevok]
Dunno but worked on second try
15:09:39 [ocharles]
you're just lucky, eventually this will break everywhere
15:09:56 [ianmcorvidae]
break all the things
15:09:58 [reosarevok]
heh
15:10:00 [ocharles]
hopefully that's fixed it
15:10:30 [ocharles]
break all the things by making everything harder than it needs to be
15:10:43 [ocharles]
cd dbmirror ; make # works everywhere except for us, because we have two postgresql's installed, and the one we really use isn't on PATH
15:10:43 [ocharles]
woooooooo
15:11:03 [ianmcorvidae]
heh
15:11:12 [ocharles]
hopefully it's fixed now anyway...
15:11:39 [ocharles]
it looks like possibly not
15:11:54 [reosarevok]
Looks like not
15:15:05 [ocharles]
still happening?
15:16:48 [nikki]
seems ok now
15:17:48 [hawke_1]
hawke_1 has joined #musicbrainz-devel
15:20:31 [reosarevok]
yeah
15:20:40 [reosarevok]
(to it seeming OK)
15:28:29 [JesseW_not_logge]
JesseW_not_logge has joined #musicbrainz-devel
15:32:31 [Freso]
Freso has joined #musicbrainz-devel
15:33:10 [ianmcorvidae]
oh good, another cycle of getting huge numbers of emails
15:33:11 [ianmcorvidae]
heh
15:33:36 [reosarevok]
heh
15:35:39 [nikki]
hopefully not another 32,000
15:37:02 [ruaok]
ruaok has joined #musicbrainz-devel
15:38:50 [ruaok]
moin!
15:39:02 [ruaok]
today hobbes will need some down-time.
15:39:10 [ruaok]
I'm going to replace one of its bad drives.
15:39:33 [ruaok]
probably within the next two hours, so plan accordigly
15:39:48 [nikki]
which one's hobbes?
15:39:58 [Freso]
ruaok: place/create is working! \o/
15:39:58 [ruaok]
test
15:40:12 [ruaok]
Freso: awesome!
15:40:19 [Freso]
Yes!
15:40:25 [Freso]
Oh, btw, thank you ocharles! :D
15:40:27 [ianmcorvidae]
test isn't up anyway, ci is the main thing that'll actually be down
15:40:33 [ruaok]
ocharles: what is the redux on how place/create was fixed?
15:41:14 [ocharles]
https://github.com/metabrainz/dbmirror/pull/1
15:41:17 [ianmcorvidae]
ruaok: summary as I understand it is place didn't have a PK set, and dbmirror was making assumptions it shouldn't have been, that resulted in an FK in that case
15:41:36 [ocharles]
i don't think it resulted in a FK
15:41:40 [ocharles]
it just seg faults
15:41:50 [ianmcorvidae]
er
15:41:53 [ianmcorvidae]
yes, segfault
15:41:56 [ianmcorvidae]
because I can type :P
15:42:00 [ianmcorvidae]
anyway, I'm off to the airport
15:42:05 [ocharles]
cheerio
15:48:25 [ruaok]
see you tomorrow, ianmcorvidae !
15:48:45 [ruaok]
I love that the ISS is in MB. :)
15:50:33 [nikki]
:D
16:02:22 [ocharles]
finally got all those failed events dispatched again
16:02:30 [ocharles]
looks like 100 of them still fail (so far)
16:02:37 [ocharles]
a lot are copies from deleted objects and stuff
16:04:44 [Sebastinas]
JonnyJD_: Sorry, haven't had the time to look more into 2.0.0-beta.5, just merged it and made it buildable, but didn't check if the patch is still necessary.
16:06:50 [ijabz]
Ive added a place :)
16:07:12 [ijabz]
but unless i use it pronto it going to go again :(
16:07:58 [ocharles]
well, you have 24 hours from the time you added it
16:08:02 [ocharles]
at least
16:09:38 [Freso]
ijabz: You can also link it up to Wikipedia or something to preserve it.
16:09:56 [Freso]
ijabz: If no Wikipedia article exists, make a Wikidata entry. :p
16:12:26 [ijabz]
ive linked it, 24 hors not that long though
16:12:41 [voiceinsideyou1]
voiceinsideyou1 has joined #musicbrainz-devel
16:13:04 [ijabz]
Alos tried to link a releasegroup to a place (recording studio) nbut you can't have to do at release level
16:14:55 [voiceinsideyou]
voiceinsideyou has joined #musicbrainz-devel
16:20:46 [reosarevok]
No, that should ideally be linked to all appropriate recordings
16:23:17 [Freso]
reosarevok: But sometimes a release simply has "Recorded in foobar & barFoo".
16:23:29 [reosarevok]
Freso: hence ideally
16:23:31 [Freso]
Oh, ideally.
16:23:34 [Freso]
Yeah.
16:23:37 [Freso]
Nvm me. :p
16:23:41 [reosarevok]
But it makes sense to keep it at the same level as recording engineer is :)
16:23:43 [reosarevok]
(IMO)
16:23:46 [Freso]
Yep.
16:25:18 [reosarevok]
* reosarevok added and linked a few places to both recordings and artists
16:25:25 [reosarevok]
Now time to order some letters :D
16:25:57 [reosarevok]
(I'm using the local printing museum as free "office" space in exchange for giving them a hand with stuff, and a lot of metal printing letters need ordering :) )
16:26:28 [andreypopp]
andreypopp has joined #musicbrainz-devel
16:34:30 [ijabz]
I would think alot of people would be interested to know that 'album x was recorded at y' more than 'song 1, song 2, song 3 was recorded at y',
16:35:56 [Mineo]
but what if you only have a compilation album with that recording? you'd have to search through all releases to find one that has the relationship
16:37:38 [ijabz]
im not really interested in compilations, that seems a less likley usecase
16:40:26 [kepstin-laptop]
kepstin-laptop has joined #musicbrainz-devel
16:45:08 [Freso]
ijabz: That's exactly what the "Display Credits at Bottom" will allow you to see.
16:45:53 [Freso]
And individual release in a release group may have "bonus tracks" that weren't recorded with the rest of the release.
16:48:57 [ijabz]
Yes i understand that, but I was thinking first of data input, if all tracks on release are recorded in same place I dont really want to add for each track individually
16:49:59 [ijabz]
and when looking at release it would be nice to see 'release recorded at', or maybe 'release mostly recorded at', I suppose that could be derived from the recordings
16:53:28 [Mineo]
using the relationship editor, adding a relationship to all recordings on a release is like 3 clicks or so
16:57:22 [Freso]
ijabz: Your first point is moot by Mineo's comment (it's one click harder to assign a relationship to all recordings on a release, than it is to do it on the release itself, if you're in the release editor). Your second point can be achieved using "Display Credits at Bottom" I already mentioned.
17:07:07 [misterswag]
misterswag has joined #musicbrainz-devel
17:15:41 [ijabz]
freso ive not display credits at bottom before thats quite neat
17:22:38 [the_metalgamer]
the_metalgamer has joined #musicbrainz-devel
17:23:14 [ijabz]
ok done it with relationship editor, I suppose if we dont have spurious multiple recordings when we should just have one it works well enough
17:43:14 [hawke]
hawke has joined #musicbrainz-devel
17:52:34 [Mineo]
ianmcorvidae: are there no places on http://musicbrainz.org/statistics/timeline/main# because there are none in the current stats or because the graph thing doesn't know about them?:)
17:54:58 [LordSputnik]
LordSputnik has joined #musicbrainz-devel
17:55:39 [reosarevok]
Stats are collected once a day, but probably both
17:55:48 [reosarevok]
* reosarevok might have forgotten that bit
17:55:56 [reosarevok]
Check the code and if it's the case file a ticket please
18:03:26 [Mineo]
I have absolutely no idea how that code works :s
18:05:38 [Freso]
Mineo: Like everyone else. So you're as suited as anybody. ;)
18:07:50 [reosarevok]
Mineo: ok, then remind me in an hour :)
18:08:18 [Mineo]
is https://bitbucket.org/metabrainz/musicbrainz-server/src/e4e202c1feba02d318d1f67f59692c2510a45194/root/statistics/view_js.tt?at=beta#cl-42 the place places should be mentioned for them to appear in the timeline?
18:21:38 [Dremora]
Dremora has joined #musicbrainz-devel
18:49:04 [kepstin-laptop]
kepstin-laptop has joined #musicbrainz-devel
18:53:59 [misterswag]
misterswag has joined #musicbrainz-devel
19:05:31 [UmkaDK]
UmkaDK has joined #musicbrainz-devel
19:12:11 [LordSputnik]
LordSputnik has joined #musicbrainz-devel
19:16:58 [reosarevok]
Mineo: at least one of them, yeah
19:17:03 [reosarevok]
So yeah, obviously forgot :p
19:17:37 [reosarevok]
Add a ticket and assign to me (so I can ask Ian for tips and fix it) or to Ian directly :p
19:21:57 [misterswag]
misterswag has joined #musicbrainz-devel
19:25:45 [Dremora]
Dremora has joined #musicbrainz-devel
19:38:08 [LordSputnik]
LordSputnik has left #musicbrainz-devel
19:43:37 [ianmcorvidae|n4]
ianmcorvidae|n4 has joined #musicbrainz-devel
19:57:10 [MBJenkins]
MBJenkins has joined #musicbrainz-devel
20:08:51 [kepstin-work]
kepstin-work has joined #musicbrainz-devel
20:14:37 [LordSputnik]
LordSputnik has joined #musicbrainz-devel
21:09:37 [reosarevok]
reosarevok has joined #musicbrainz-devel
21:37:37 [Freso]
Any ETA on getting test.mb up again?
21:38:06 [Freso]
Or I guess I could try and get my rika install running again.
21:39:28 [nikki]
I think everyone's too busy fighting fires to set up a server
21:39:40 [nikki]
I mean, it's not like we were making any progress *before*
22:06:41 [kepstin-work]
kepstin-work has left #musicbrainz-devel
22:14:13 [ruaok]
ruaok has joined #musicbrainz-devel
22:15:20 [Freso]
How much space does the pgsql db use these days?
22:16:34 [reosarevok]
A couple cubic light years
22:17:36 [nikki]
under 20 gb
22:17:59 [nikki]
under 17 gb it seems
22:18:06 [nikki]
given that my server has 17 gb used :P
22:20:49 [Freso]
* Freso looks at his 4.1 GiB free space on the partition that has the pgsql data dir...
22:20:52 [Freso]
:/
22:29:45 [ruaok]
ruaok has joined #musicbrainz-devel
22:30:37 [ianmcorvidae]
'sup
22:30:40 [ianmcorvidae]
* ianmcorvidae is in San Jose
22:30:47 [ianmcorvidae]
ruaok is popping in and out, it seems :P
22:35:15 [ruaok]
ruaok has joined #musicbrainz-devel
22:36:15 [Freso]
That's what s... nvm.
22:54:12 [misterswag]
misterswag has joined #musicbrainz-devel
23:09:36 [LordSputnik]
LordSputnik has left #musicbrainz-devel
23:16:54 [Dremora]
Dremora has joined #musicbrainz-devel
23:27:50 [misterswag]
misterswag has joined #musicbrainz-devel
23:49:43 [JonnyJD]
JonnyJD has joined #musicbrainz-devel