legoktm: not yet, no. but we're working hard to change that.
2015-12-21 35534, 2015
Freso
legoktm: No, but they're there.
2015-12-21 35542, 2015
bitmap
I was wondering if we could schedule downtime to upgrade to pg 9.4 soon (but obviously not use any >9.1 features until May or so)
2015-12-21 35545, 2015
chirlu`
Also, we recently moved the chatlogs; it would be strange to move them again soon.
2015-12-21 35508, 2015
bitmap
we have some pretty bad performance issues with edit listings that I think index-only scans and other improvements will help a lot with
2015-12-21 35509, 2015
ruaok
we can dual home them.
2015-12-21 35514, 2015
ruaok
and then change later.
2015-12-21 35523, 2015
ruaok
but in general I agree with you chirlu`
2015-12-21 35546, 2015
zas
bitmap: this is where we could put baron in play
2015-12-21 35504, 2015
reosarevok
bitmap: and we could use those index-only scans already now and not from May?
2015-12-21 35509, 2015
legoktm
As someone who spends a lot of time trying to explain the difference between Wikipedia/Wikimedia and the confusing naming, I don't think it's worth giving up the already decent brand name of MusicBrainz for a significantly lesser-known one. Also https://www.google.com/trends/explore#q=MusicBrai…
2015-12-21 35533, 2015
bitmap
reosarevok: yep, it's a backend thing that we'll benefit from automatically
2015-12-21 35501, 2015
ruaok
legoktm: we're not giving up on Musicbrainz in any capacity.
2015-12-21 35516, 2015
ruaok
it makes less sense to run something like BookBrainz under the MusicBrainz umbrella.
2015-12-21 35529, 2015
bitmap
zas: so use baron as a read-only 9.4 slave, perhaps?
2015-12-21 35547, 2015
zas
bitmap: yes something like that
2015-12-21 35500, 2015
zas
i can prepare the server for that
2015-12-21 35526, 2015
zas
then at some point we set totoro as RO and baron as RW
2015-12-21 35549, 2015
chirlu`
Can Postgres servers of different version do log-shipping replication?
2015-12-21 35556, 2015
chirlu`
I seem to remember that’s not possible.
2015-12-21 35559, 2015
bitmap
I think slony can
2015-12-21 35504, 2015
ruaok
chirlu`: good question.
2015-12-21 35509, 2015
ruaok
ugh, not slony no.
2015-12-21 35517, 2015
ruaok
I'm ok with some DB upgrade downtime.
2015-12-21 35528, 2015
ruaok
but getting baron up for read-only would allows us to stay up.
2015-12-21 35529, 2015
ruaok
limping, but up.
2015-12-21 35541, 2015
bitmap
cool
2015-12-21 35558, 2015
bitmap
any ideas for when?
2015-12-21 35559, 2015
ruaok
we should finally drop the pre-NGS tables.
2015-12-21 35513, 2015
ruaok
doing the upgrade without that is nearly impossible.
2015-12-21 35533, 2015
ruaok
in the new year? not sure how people's attention will be before then.
2015-12-21 35540, 2015
bitmap
alright, I need to start a google doc with upgrade notes
2015-12-21 35543, 2015
ruaok
certainly after xmas?
2015-12-21 35544, 2015
reosarevok
Which tables are those?
2015-12-21 35556, 2015
ruaok
reosarevok: tables not in any schema.
2015-12-21 35505, 2015
reosarevok
Oh
2015-12-21 35507, 2015
reosarevok
Ok
2015-12-21 35509, 2015
ruaok
we have the complete pre-NGS data still sitting in the public schema.
2015-12-21 35529, 2015
ruaok
this is something that will need zas' attention.
2015-12-21 35543, 2015
ruaok
and the caveat about search servers applies here as well.
2015-12-21 35546, 2015
ruaok
at least in my mind.
2015-12-21 35554, 2015
reosarevok
That being?
2015-12-21 35527, 2015
stanislas
LordSputnik: ping
2015-12-21 35544, 2015
ruaok
that the search servers suck and still take a lot of time.
2015-12-21 35549, 2015
reosarevok
stanislas: wait 10 min until the end of the meeting please :)
2015-12-21 35552, 2015
ruaok
we should put all available effort to fixing those.
2015-12-21 35559, 2015
Freso
stanislas: Please wait 10 minutes. :)
2015-12-21 35509, 2015
reosarevok
Well sure, but if we need java people for that, I dunno how much zas can do
2015-12-21 35509, 2015
opatel99
Gentlecat: If the url is /release/ in CritiqueBrainz, then isnt it easier to jump from CB to MB and other
2015-12-21 35527, 2015
reosarevok
opatel99: 10 min please you too :)
2015-12-21 35535, 2015
stanislas
ok
2015-12-21 35538, 2015
ruaok
let's do this.
2015-12-21 35553, 2015
ruaok
bitmap: you start the doc. zas: you keep an eye on the doc and help out a bit.
2015-12-21 35502, 2015
zas
ok
2015-12-21 35504, 2015
ruaok
then in the next few days lets discuss this in more detail.
2015-12-21 35509, 2015
ruaok
and compare schedules.
2015-12-21 35523, 2015
ruaok
onward, Freso
2015-12-21 35523, 2015
bitmap
sounds good
2015-12-21 35525, 2015
TOPIC: MetaBrainz Community and Development channel | MusicBrainz non-development: #musicbrainz | MeB meeting agenda: Reviews; requirements for CAA-84 (reo), CB repos (Gentlecat)
2015-12-21 35532, 2015
Freso
reosarevok: CAA-84, go. :)
2015-12-21 35543, 2015
zas
ruaok: power up baron, so i can make it ready and configured with usual stuff
2015-12-21 35559, 2015
reosarevok
My trello notes for this say
2015-12-21 35501, 2015
reosarevok
"The IA want to know if we can use the same bucket as for cover art (IMO no), and whether they are treated differently and should be derived differently (are original, 250, 500 a reasonable choice for posters too?). We need to answer this before they can look into setting it up. If we want different derivation, we should try manually and decide what sizes we
2015-12-21 35501, 2015
reosarevok
want.""
2015-12-21 35517, 2015
reosarevok
(this being the event poster etc archive)
2015-12-21 35531, 2015
Freso
I agree with using a different bucket. Can we start by using the same derivations and ask them to change those later once we've had a chance to play with it?
2015-12-21 35558, 2015
Freso
(I mean, we'll probably want larger than cover art, but how much? No clue.)
2015-12-21 35540, 2015
reosarevok
I suspect this being the IA, it won't happen tomorrow, so we could play with it
2015-12-21 35545, 2015
reosarevok
GCI research? :p
2015-12-21 35553, 2015
Freso
Gopher it.
2015-12-21 35533, 2015
ruaok notes the request to DWNI has been sent
2015-12-21 35550, 2015
chirlu`
For gopher.metabrainz.org?
2015-12-21 35556, 2015
Freso
:D
2015-12-21 35505, 2015
zas
:)
2015-12-21 35533, 2015
reosarevok
Gopher would be neat :p
2015-12-21 35551, 2015
Freso
Any other comments? Do we have time for Gentlecat's topic?
2015-12-21 35504, 2015
Sput|Away
reosarevok: would it be possible for you to mention about book cover art when you next email IA, if it's appropriate to slip it in anywhere?
2015-12-21 35512, 2015
reosarevok
But ok, different bucket, similar code, maybe different sizes but we need to check it
2015-12-21 35530, 2015
reosarevok
Sput|Away: heh. Sure I guess. I wonder if Brewster has heard of BB
I think we talked about it a bit 3 years ago when Brewster was at the summit?
2015-12-21 35554, 2015
Gentlecat
and do code review there
2015-12-21 35504, 2015
Freso
Obviously, BB was very much VapourWare back then.
2015-12-21 35514, 2015
ruaok
Gentlecat: fine by me
2015-12-21 35515, 2015
Gentlecat
people already submit pull requests on github, and we moved MBS reviews there recently
2015-12-21 35517, 2015
Freso
Gentlecat: +1 from me.
2015-12-21 35522, 2015
Sput|Away is now known as LordSputnik
2015-12-21 35529, 2015
ruaok nods
2015-12-21 35530, 2015
bitmap
+1
2015-12-21 35532, 2015
Gentlecat
alright
2015-12-21 35544, 2015
Freso
I was actually surprised when someone wrote here that the docs said to use BBucket for that. :p
2015-12-21 35546, 2015
Gentlecat
will start baking then
2015-12-21 35551, 2015
Freso
Alright!
2015-12-21 35556, 2015
Freso
Thanks for your time peeps!
2015-12-21 35557, 2015
LordSputnik
reosarevok: thanks, let me know if they say anything positive please :)
2015-12-21 35500, 2015
Freso
</BANG #2>
2015-12-21 35508, 2015
ruaok
thanks freso
2015-12-21 35524, 2015
Gentlecat
Freso: well, that's how it was supposed to work
2015-12-21 35531, 2015
Freso goes to catch up on GCI for the 2nd time today :p
2015-12-21 35544, 2015
Freso
Gentlecat: Sure, but I didn't realised it hadn't moved back when MBS did. :)
2015-12-21 35559, 2015
Freso
*didn't realise/hadn't realised (your pick)
2015-12-21 35526, 2015
Gentlecat
opatel99: I'm not sure what you mean by "jump". if you want to show release group info on /release/<mbid> page, that's not a right thing to do
2015-12-21 35543, 2015
Gentlecat
it needs to lookup release group in MB and redirect to its page
2015-12-21 35559, 2015
opatel99
Ok
2015-12-21 35510, 2015
Gentlecat
Freso: I pick third option
2015-12-21 35514, 2015
opatel99
Gentlecat: So given a /release/x url, we are to lookup x for its release-group and redirect to /release-group/x.releaseGroup()
2015-12-21 35542, 2015
Gentlecat
precisely
2015-12-21 35506, 2015
Freso
YatharthROCK_: Yeah, unfortunately we're a bit noobish in the GCI department, so we didn't have a way really to mark tickets that had been entered as tasks, so it required some manual overview to keep tasks in sync with the couple of tickets that were fixed out of GCI, and apparently one managed to slip through our fingers.
2015-12-21 35541, 2015
Freso
YatharthROCK_: Feel free to make the PR and comment about it on Jira and Gentlecat/bitmap can decide which approach is best.
I won't promise that we'll re-open the task for you, but I won't promise we won't do it either. :)
2015-12-21 35520, 2015
stanislas
LordSputnik: I wrote some time ago about current tests for relationships and that they are not usable anymore (they just use some constants). I am currently in the final stage of repairing test_relationship.py. I will make some commit, containing the changes in about 10 minutes. What do you think ? Can I change this relationships file
2015-12-21 35544, 2015
LordSputnik
stanislas: why is it broken though?
2015-12-21 35515, 2015
LordSputnik
Is there no way to keep the existing fixtures.py data and add the new data on top of it to prevent test_relationships.py from breaking?
2015-12-21 35505, 2015
stanislas
I am rewriting relationships.py to get the data from the database, it would be ok then.
2015-12-21 35520, 2015
stanislas
It will test the same data and have the same functions.
2015-12-21 35545, 2015
gcilou_
opatel99, btw the code in account.pm will not work :(
2015-12-21 35556, 2015
stanislas
LordSputnik: Just not using constants from code and from database
2015-12-21 35516, 2015
stanislas
It won't break then.
2015-12-21 35511, 2015
LordSputnik
stanislas: OK, let me know when it's done and we can finalise things
2015-12-21 35515, 2015
LordSputnik
Thank you
2015-12-21 35528, 2015
opatel99
gcilou_: What is wrong?
2015-12-21 35520, 2015
gcilou_
opatel99: apparently there is no such parameter for userid so the statement $c->request->params->{userid} does nothing. The info about user is in the $editor
[20:02] reosarevok Gentlecat: so do I and I live in Estonia... it's like +11 here today, wtf
2015-12-21 35525, 2015
CatCat
it's fukken like 10 here too, wf, december i norway. wtf
2015-12-21 35530, 2015
gcilou_
reosarevok: So you were right, the username needs to be accessed by $editor ...is the information about $editor in account.pm?
2015-12-21 35556, 2015
ruaok sees that we received a cease and desist message from another lawyer who doesn't understand what we do.
2015-12-21 35558, 2015
ruaok
fun fun fun.
2015-12-21 35509, 2015
ruaok closes laptop for real now.
2015-12-21 35510, 2015
reosarevok
gcilou_: $editor should be just a specific user, so try just $editor->name maybe?
2015-12-21 35508, 2015
gcilou_
reosarevok: Ok.. so there's like this piece of code in account.pm "$c->model('Editor')->update_email($editor, $email); if ($c->user_exists) { $c->user->email($editor->email); $c->user->email_confirmation_date($editor->email_confirmation_date); $c->persist_user(); }" does it have anything to do with what we are doing?
2015-12-21 35527, 2015
CatCat also thinks that chatlogs, wiki, blog, etc should stay on *.metabrainz
