#musicbrainz-devel

Does the musicbrainz server virtualbox image include the web service? If so, does the web service have the rate limiter?

it includes the webservice but the ratelimiter isn't included in the VM image

in fact the ratelimiter is, thus far, still private

Excellent. :)

Thanks ianmcorvidae.

good luck :) hope it works for you

Thanks. :)

note that the default setup you'll get with the vm is actually the development server, which isn't exactly great for performance.

the VM at all is unlikely to get great performance, it's a tiny virtualbox image :P

A local server without rate limiting will still be faster than querying the official webservice, though, so I'm okay with all that.

it's quite sufficient for one person, just don't deploy it for your whole office using picard or anything

Understood. :)

ianmcorvidae: Well, that depends on the size of the office, I'd say...

bah, now I'm looking at implementing the Maybe monad in python

I haven't even written that much Haskell! Really!

eh, I guess I probably don't need it

I just hate try/except :/

warp: Can you take on OTHER-133 like last year

navap: if he can't I can

It's a race to click the 'assign' button!

(obviously in two days or so :P)

heh

Did you ever get time to condense the summit notes a bit?

And by a bit I mean a lot :)

ocharles basically did that as his blog post

that's a reasonable summary, so I didn't ever condense them myself :)

Okay

Going through the blog posts for 2012 is going to be a pain. So many weekly server update posts :/

heh

luks: hmm, could you update http://users.musicbrainz.org/~luks/picard-plugi... to support the unicode apostrophe too? i.e. ’

nikki: that's the plugin from https://github.com/musicbrainz/picard/blob/mast... so you can do that, too!:)

well, I just looked at what's on the plugins page

since presumably that's where people get plugins from...

navap: yes.

Beta is very slow for me right now.

https://gist.github.com/f45599dad255740d5bab -- the NES merge stuff can now render as HTML or JSON

constant applicative functors + monoids + semigroups & reducers = AWESOME

Though I'm still not sure about the whole [Hunk monoid] thing, so we're still not quite at 100% yet

luks: btw, if you do change it, could you leave a note on http://forums.musicbrainz.org/viewtopic.php?pid... ?

ianmcorvidae: You need to put some time aside for reviewing the OAuth patch. I don't want to give Geordi my password every time I go to it. :p

Freso: huh? I have never had to re-login

I've had to a few times

Curious :)

The three times I've visited, I've had to login.

reosarevok: Do you also not have to log in to AcoustID?

Dunno, I rarely visit the site

But I don't remember re-logging in there either

(when I have)

ianmcorvidae: How do you import something from Geordi to MusicBrainz? Or is that not available yet?

not available yet

Aw.

I've just copied and pasted the tracklist a couple times

Not that much to remove, just the acoustID stuff

But yeah, an import button would be nice :)

luks: ah, thanks :D

Speaking of Picard...

ianmcorvidae: Ping?

he's probably still asleep

Yep. And now there's a ping waiting for him when he wakes up. :D

Freso: pong

Freso: there's no import function in geordi yet, just matching

Freso: though obviously it's on the roadmap :)

I guess import would be release editor seeding?

that is one sort of import, yes

for what's mapped at present I think that's the only one that'd be needed, but ultimately there will presumably be some non-release stuff :)

notably I know of some stuff that will have URLs

Freso, nikki, reosarevok: your needing to relogin or not is probably a product of your respective browser usage patterns and preferences; I've never had to relogin either (except when I explicitly log out), but I also basically never close my browser :P

I don't want to have to worry about whether I'm going to be logged out if I restart my browser :/

what i really hate is when a site logs you out in your other browsers when you log in with a new one

oh yeah, that's evil

ianmcorvidae: How would one go about getting more datasets added to Geordi?

ianmcorvidae: I'm thinking of http://smdb.kb.se/ in particular right now. Should the "dataset" be contacted? Should you/MB contact them? Are we "on hold" waiting for further Geordi/WCD development before looking into getting/adding more?

ianmcorvidae: Do you want me to ask you more questions, or do you want to answer those first? :D

I think the answer to all of those is "we don't know"

however, personally I would love for there to be more datasets

ianmcorvidae: your last email to mb-devel (and a few more) seems broken with half the headers in the message

I've been looking into adding Jamendo, vaguely

bah, probably my client broke it again :/

http://smdb.kb.se/ looks really comprehensive for Swedish stuff. Not that I have a deep interest in Swedish, but if it'd allow us to add composers and lyricists and performers to a bunch of (Swedish) stuff... :D

most of ian's emails seem to be like that :P

it sometimes adds linebreaks in the subject

"most" is a stretch :P

hmm. I should extract some data from the stuff I've acquired

that's not a terrible start

if you can get a dump out of what they have, and reason for me to think it's fine to import it, I can do that :)

hmm. would it be possible/make sense to somehow add things to geordi where you can't get a dump, but just have access to e.g. a search and screen-scraping?

the import scripts might need tweaking, depending how their stuff is formatted -- right now it's optimized for "a bunch of directories full of XML + JSON files, with the directories named according to the identifier they should be under in geordi"

well, search + screen-scraping is a way of acquiring data :P

I think what's unclear, at least to me, is what the scope of "things we want in geordi" really is

derwin has been doing things like extracting metadata from large collections of scene-released music among his friends

which I'm unclear if that's what we want

but on the other end things like Jamendo, Discogs, etc. seem straightforwardly useful

(and have explicitly public-domain data)

stuff like this smdb.kb.se thing, stuff we have to scrape, etc., get more fuzzy

however for now geordi is very beta

so give me all the datas :D

(especially if you'll subsequently send me pull requests to develop the mapping for said datas)

can we talk about having the bot know when meeting holidays are?

kepstin: hah, that is certainly evil.

it seems ridiculous to have an automated email system that can't be configured for, you know, emailing on the right day

heh, the bot thing is really silly

presumably I should have prodded rob to figure out where it's running so we could just stop the cronjob for a couple weeks

(assuming it's running within MB, I'm not sure)

yeaah

(see its name :p)

I just found the source. maybe we can fix it

IIRC ruaok built it one day because he was tired of everyone asking when the meetings were

I think "one day" may be overestimating the time expenditure

so, it seems useful to have something that says when the meeting is then :)

I submitted a patch to make the UUID formatting closer to what MB does ;)

let me look at it this week, maybe

I thought the problem was more remembering to send emails

so people never knew if the missing email meant no meeting or just a forgotten reminder

whereas now there's a bot sending them and people are like "wait a minute, that's wrong *reply*"

yeah, that sounds right

Oh, ok

and sometimes people wouldn't make new threads, which I think is the reason for the "issue <UUID>" thing :P

ianmcorvidae: your reply today isn't threaded with the original mail. odd.

(in my client)

because his headers are messed up

my client screwed it up

because vim ignores my settings in some weird way I've never deciphered

the beauty of open source is that if you dont like it, you can fix it. :)

ianmcorvidae: Well, I'll try e-mailing SMDB then, and see what they say. They're freely giving the data to researchers, and as far as I can tell, the data collection is under the Swedish state. I have no clue what the Swedish state currently thinks about Open Data, but I know the Danish one recently releases a whole bunch of data into the open.

cool

ruaok: Have we made any changes to our server racks since the photo you took in Jan 2012?

the one person I tried emailing hasn't gotten back to me, heh