#musicbrainz-devel

ianmcorvidae: I think I can start working on UI improvements and documentation now. And then probably sqlalchemy integration and api, after getting a better idea about how stuff works.

unless there's anything more important

can you share a bit more info about what needs to be done there?

Gentlecat: that sounds like a good place to start, to me

the biggest outright missing features are matching, importing to MB, and more data sources, in that order of priority

matching I've stubbed out a bit but it may not be clear what I intend from what I've done; importing to MB/seeding I'm guessing you've never worked with on the MB side either

more data sources can be worked on but it's mostly writing the mapping stuff, so I'd say making the UI more clear, documentation, etc. are higher priority

is that what you wanted to know, or was there something more specific re: matching, API, sqlalchemy, etc.?

what's the difference between old version (on http://geordi.musicbrainz.org/) and the new one? not sure if you talked about it

well, it's completely rewritten :)

the change that precipitated the rewrite was moving the data storage from elasticsearch (wasteful, slow, bad scalability) to postgresql (awesome)

but also making it much easier to write mappings

the old version just had a big monolithic "map" function that did it all

and setting it up to be a bit better for further work; the old version it would be impossible to implement something like merging, and it's harder to move in the direction of materializing matches from the MB side, like the discogs stuff I was talking about

and I guess basic browsing that is currently implemented is just a temporary replacement for searching

well, the browsing stuff will probably stay, but yeah, search does need to be added

notion there was probably to use solr in some capacity, but I haven't really done the research on that yet

(which shouldn't block others from working on it, I just mean the technical bits of how to wire it all together, not any design decision type stuff)

huh, I didn't know about minus sign feature in jinja templates

should start using that in CB too, we've got a lot of whitespace there :)

:)

in MBS's template system there's four levels of it, we mostly only use the - but we've started getting a few ~ and = in there

I forget exactly which one jinja's {%- -%} corresponds to though

TT is a really weird template system :P

is it Perl specific thing?

TT is perl, yeah

it's not any more a perl thing than jinja's a python thing, neither's a cross-language template system, but yeah

I never worked with it, so no idea how it works there

yeah -- TT has the one that trims nothing (obviously), - trims whitespace up to and including one newline (but not further), = cuts all whitespace (including many newlines etc.) into one space, and ~ removes all whitespace before whatever's next

is there any reason why you wouldn't want to cut all whitespace?

well, sometimes you want the space there

since it's not always just between HTML tags or whatever

if you have [% entity.name %] ([% entity.comment %]) you don't want it removing the space before the (

TT doesn't distingish between what jinja does with {% and {{

okay

write once, present often. another version of Alex Gaynor's code review presentation (if anybody cares): https://www.youtube.com/watch?v=mhU3xwPGJ7I

Picard developers: please release Picard 1.3. More and more, people are coming to the forums with questions and problems with 1.2 that are solved by telling them to download and install the 1.3 daily build. Also: Picard 1.2 was released more than a year ago while there have been massive improvements to the program since then- it’s time.

thank bitmap for the hint, zastai i added this MAX_ENTITY feature in my super turbo script

blup too↗  :)

(upploaded to github should update mirrors openuserjs and greasyfork)

ianmcorvidae: should I just fork and submit pull requests from it?

yeah, that's probably easiest/best

I haven't imported any data yet, but http://geordi.musicbrainz.org/ is also the new code now

oh, I was actually using old one as a reference :)

oh, heh

but code is there, so that's not a problem

ianmcorvidae: can we set up a sandbox for code reviews?

we can do that later if you don't have time right now

I'll submit PR tomorrow anyway

yeah, we can do that, but I'm off to bed right now

yep

bah, no gentlecat

ruaok: do you know what happens in cb if artists or release groups get merged or deleted in mb?

currently nothing.

there is no active link of any kind between MB and CB.

besides CB being a client of the MB API

so it stores the artist locally?

lemme review the schema

looms to be dynamically retrieved: http://critiquebrainz.readthedocs.org/db.html

looms? lookes even. we're not weavers.

so if the release group got deleted, I guess it would fail to look it up and then the page wouldn't display right

likely. worth trying out.

I was also going to make a ticket for saving drafts, preferably automatically

nice idea too.

and I haven't written any reviews, but I suspect after publishing one, there's no "share this on facebook/twitter/etc" thing

(unless there is, in which case I won't make a ticket for it :P)

there isn't, but should be

k

and gentlecat has plenty of time left. :) :)

he probably doesn't even *need* plenty of time with the speed he works :P

srsly.

another thing that came to my mind regarding the switch to https: the CAAs json files contain links to releases/releases groups with http, should we regenerate those or at least insert https links in new files?

ideally both.

definitely the latter, that shouldn't be hard

probably just an 's' in https://github.com/metabrainz/CAA-indexer/blob/...

I like how the GSoC organizers can make it uttlerly clear what the deadlines are. I'm looking at one email where the deadline is shown 5 times. 3 times in bold.

I bet people are still going to "miss" it.

at least there is not chance in hell of someone claiming it wasn't clearly visible

heh. paypal's new make-over looks suspciously much like WePay.

at least they have some nice new map view of where your donations come from. useful for when I get around to the annual report.

ok, finances are done. Took 2 hours, but at least I have peace from quickcrooks for a month

interesting: http://www.billboard.com/biz/articles/news/publ...

New post: blog: New autumn schema change date: 17 November, 2014 <http://blog.musicbrainz.org/2014/07/01/new-autu...;

Hmm, interesting that the (release) relationship editor fills in the recording name rather than the track name when doing 'add related work'. I can’t decide if that’s a problem or not. :-/

ocharles (and/or others): is the code at https://github.com/metabrainz/CAA-indexer up to date? the tests fail (https://gist.github.com/mineo/f9f60d0c045fa598cc08) although I didn't touch a thing :/ fixing most of it seems easy but I'm wondering if I just got the wrong branch or something

ianmcorvidae ping

Mineo: looks like it might be a package versions thing, not sure

I've gotten past the rabbitmq message now and opened https://github.com/metabrainz/CAA-indexer/pull/12

ah, heh

I'm not surprised the tests weren't updated, then :P