#musicbrainz-devel

it is

so one sec and you can try again

dunno if it'd need some sort of change to make it use the py2.7 dev headers

thanks

that's installed if you want to try again, anyway

yes, it worked

yay

ianmcorvidae: http://gentlecat-geordi.mbsandbox.org/

hm

I guess psycopg2 defaults to TCP auth

use 127.0.0.1 rather than localhost, since localhost defaults to ipv6 localhost which oidentd doesn't listen on

oh jeez, that again

the number of times it's caused me problems, and I don't even use python much :P

heh

the first time it came up it was java IIRC

it's always been python for me, afaik

anyway, yeah, add host=127.0.0.1 to the connection bit I Think

think*

hm, though it claims it defaults to unix socket if not provided

weird

it worked, I'm trying to compile styling

alright

npm install from the top directory and then it'll be in node_modules/less/bin/lessc

though you have to define input/output files directly right now, no script for that

but it's just turn geordi.less into geordi.css anyway

hm

too old a less version, looks like?

it worked fine with JS

maybe

yeah, needs 1.5.0 I guess

which I should fix in the dependencies, but

and yeah, the less.js one is from 1.7.something I think

1.7.1 I guess

(sorry to interrupt your CB discussions) this is my daily plea to the Picard devs to release 1.3. evidence that this is a good thing: http://forums.musicbrainz.org/viewtopic.php?id=... people want to use Picard, but 1.2 is vastly out of date compared to 1.3. most people don’t even know where to find the daily builds. please finalize anything that needs to be done and release 1.3.

ianmcorvidae: I don't think I can update it

can you help me with that?

Gentlecat: I don't see a node_modules folder in the geordi folder at all, where do you have it installed?

I didn't install anything node-related

lessc is a nodejs app

I was using lessc that is installed right now on rika

I'm not sure where you're getting lessc from is what I'm wondering

hm

yeah, I know

hm

yeah, I think installing it local to the directory would be best

from the top directory of the geordi checkout, 'npm install'

and then you'll have a lessc in node_modules/less/bin/lessc

nikki: who’s in charge of Picard? like I said yesterday, more and more over the past few weeks, people are coming to the forums with questions and problems that are resolved by telling them to obtain the daily builds of Picard. It’s time for 1.3 to be released. Who would be the person to bug to get this to happen?

the answer is not to bug people :(

ianmcorvidae: http://gentlecat-geordi.mbsandbox.org/data \o/

haha whoo, you reintegrated museo slab

anyway yay!

I created PR. will go eat something now :)

I'm not sure who's in charge any more. luks used to do the windows builds, but afaik he's looking for someone else to take it over

thanks for helping

but yes, as CatQuest said, the answer isn't to bug people

hm, logout seems not to work? weird

or the front page just won't interact properly with my cache or something, bah

works fine here

if people are never bugged, they will never see this is a problem. people who are new to MB and all of its goodness will continue to download Picard 1.2 and continue to have problems and continue using an inferior product. 1.2 doesn’t have covert art download capibilities (except through a plugin that barely works). 1.2 automatically tags any release with more than one artist as a compilation, which is wrong. 1.2 cras

or won’t even load for some Windows 8 users (i’m not sure of the conditions) while 1.3 works just find for them. some of the devs, like bitmap, have stated they rarely read the forums, so they do not know about these issues. i’m championing for those on the forums whose voices are not heard.

I can get the login button to reappear on other pages, but the front page will only show me logged in

ianmcorvidae: custom-variables.less was a replacements for variables.less, right?

yeah, it was variables.less with all the paddings/margins cut in half (basically)

I thought that this file came from bootstrap and just replaced it

yeah

I'm not sure if there's a better way to do it just overriding the things we care about

and now logout works for whatever reason

weird

anyway seems fine, I just wanted to look at the login-button styling again to make sure I liked it :)

does checkbox look good?

yeah -- I think I liked the slide-out a bit better, but the popup isn't bad

is the popup thing provided by bootstrap.js, or?

yes, that's bootstrap.js

it's also used to collapse navbar

try changing window width

there's a small issue with checkbox block

not sure how to fix this

the width thing when it's collapsed, I guess?

yes

nikki: I created a PR for the iTunes importer: https://bitbucket.org/96187/userscripts/pull-re...

thanks

np :)

KillDaBOB: we're already aware that a new release is long overdue. as I said, the people who work on picard are volunteers, they're doing it in their own time for nothing. coming here and telling people to do a release and repeatedly talking about "bugging" people doesn't come across well, and isn't likely to make people want to have anything to do with it

I mean, who would want to admit to working on picard when you've just said you want to bug them daily until they do a release? :P

ianmcorvidae: tooltip stops working after I move JS to the bottom of the page

yeah, I noticed that was an inline script

the inline script can presumably be put in block js to move it down though

oh, right

ianmcorvidae: do you want me to add inline comments there https://github.com/metabrainz/geordi/pull/13#di...

it probably doesn't need to be in the code itself, I just don't understand bootstrap terribly well so I wanted to confirm I understood the right thing (i.e., when it's under a certain breakpoint, put it to the right (not sure exactly what pull-right does, but) and pull it to width 100%)

okay, I'll add a comment in PR

ianmcorvidae: I decreased the gap between navbar and h1

is it better now?

yeah, that looks better

nikki: point taken. my voice has been heard. a lot (not all) of the people who contribute to Picard hang out in this channel, so my hope is that they will read their logs and see this. i’ll keep quiet- for now.

heh, I guess I should write matching soon so the login actually has a purpose

I was thinking about trying to migrate to sqlalchemy

to make code cleaner and development easier in general

probably would be good, would make it easier to do the API stuff too

yeah

I just went more with what I know :)

“Ned g’schimpft is g’lobt g’nug!”

(re: pep talk)

ianmcorvidae: are issues here http://tickets.musicbrainz.org/browse/GEORDI from pre-rewrite period?

are they still relevant?

Mineo: I understand it as “nicht”.

fair enough :)

Gentlecat: they're from the pre-rewrite period, though some of them might be applicable still, or solved by the rewrite

anything data-source-specific is certainly not applicable until we add those data sources back

GEORDI-17 is half-solved by the test dataset, GEORDI-20 is definitely still applicable

the stuff about importing subitems (no longer a concept that exists in the new version) and other entities are things the rewrite's supposed to help with, but there's no importer yet

the automatic matching stuff is a can of worms we haven't really opened with the new code, though the intention is to move them into the geordi codebase rather than having them be external things talking to an API (which should help)

I think a lot of the rest is data source specific

okay

lol the ninjatune data

has 'NOT FOR RELEASE OR EXPLOITATION' in a barcode field somewhere, it appears

bitmap: updated that release events/geordi PR

cool

if you have a better idea for the no_manip thing I'm all ears, heh, that was definitely a hack

but it's not hopelessly repetitive anymore, which means even though I didn't do named functions it's probably not as insane to parse all the lambdas and whatever any more :P

probably

it's definitely easier to read now

where in the code is the PathPart normalization process you mentioned in the one comment?

https://github.com/metabrainz/geordi/blob/maste... is where it's called

which calls https://github.com/metabrainz/geordi/blob/maste...

ah

which initializes a PathInserter class with the path as provided and then uses the source_value() of each path part it creates to go back

SimplePathPart's source_value is what uses no_manip, around https://github.com/metabrainz/geordi/blob/maste...

and the no_manip means the last source_value remains a pathpart

it means that whichever one it's set on remains a SimplePathPart, yeah

which means the code at https://github.com/metabrainz/geordi/blob/maste... won't append an OrderedPathPart() instance to the end

as mentioned I'm pretty sure there's a better way to be doing this, but I don't want to force every path to stick "('some fake name',)" at the end or whatever

oh, I guess it'd just be (,) but that's still silly

one option is to stop accepting non-array things, i.e. let it be possible for a single release event thing to list more than one date and more than one country

but that seems silly given you can/should just add another release event thing instead

it'd be useful if you don't have the connection between dates and countries, but there's a separate place in the data format for so-called split data like that

nn!

aw, he left :P

I think I'm confusing tuples in the source path and tuples in the destination path