#musicbrainz

Maybe it'd be better left as just a basic AR, rather than attribute, "miscredited"

not to mention that it could be done that if some "x is mmiscredited to y link is in place, any y having a "x" added under it would return an error

CatCat: Is that “I saw it on a shitty discography site once” or “someone posted about it on twitter” or “it’s in the liner notes wrong”?

I can’t fix the conflict between paste-a-date and remember-dates because your script is using innerHTML ← this is what is bugging everything related to DOM manips

you want that link, but you don't want to spam a work with lots of ARs saying someone *didn't* do something :P

BrianFreud: right, exactly.

there have been releases that show both that, and a weird al songs wrong

they were bootlegs.. but still

BrianFreud: The other problem is that that information doesn’t show up anywhere that it would help.

Probably the one place it would, would be in the comment

BrianFreud: it’d be interesting for there to be a comment for each alias

hawke_: like in the case of rabbit join "all over the f* place"

yeah; I find the current "locale" attrib for work aliases isn't really useful

BrianFreud: so for example if you searched for babylon and got a result of Belchezaar, it could tell you why.

in the case of weird al (and the rabbit joint too actually) on an actual physicak bootleg release

basically "overvelmingly so"

So you’d know it wasn’t just the shitty search engine finding something useless. ;-)

CatCat: I know, but that doesn’t help for when to use an AR.

yeah, http://tickets.musicbrainz.org/browse/MBS-4016 would go partway to that, but even then you don't know why Babylon is an alias to Belchezaa

i'd say definitely in cases where it is a) more often known wrongly, b) known as /because/ it is "miscredited"

as those two examples would testify

well at least the zelda thing

BrianFreud: "locale" is not too useful *yet*

Because we're not doing anything with it :p

reosarevok: locale is kind of questionable anyway, as others have pointed out it should be script or something.

hawke_: well that's what the voting system is for ;)

hawke_1: locale / script both work for me

I'd be happier with something useful; "mentioned in a book", "artist's website", "some liners", "ASCAP db", "ISWC db", "SESAC db", "copyright registration title", etc

reosarevok: locale alone doesn’t cover different scripts though. romaji, katakana, hiragana, kanji can all be “japanese”

" it will be useful

Fuck the chat and line breaks

hawke_1: probably

aliases need script

yes, it's useful for trans*tion, but not for IDing what kind of alias an alias actually is, or where it's from.

jesus2099: locale is pretty useless for artists

jesus2099: BEATLES is not a word. :-p

Not so for works

YAMAGUCHI Momoe then, always japanese

(for artists, it's mostly useful for a) orchestras and b) legal bullshit)

jesus2099: Therefore not an englsih word. ;-)

and not t omention situations wher it *looks*english, but it isn't

hawke_1: but we have to use the english accent to it, so it is english proper noun

I do agree with adding script to them somehow though, so you don't have to convince me :)

reosarevok: Just that it makes it even more complicated. :-)

jesus2099: hehhehhe do we? You can't imagine the amount of Spanish people who say it in Spanish

for example this TNT song ;:

Not that much for the Beatles maybe, but ask about, dunno, Oasis or Placebo

(np

“Los Beatles”?

Now grooving to: TNT ~ Klassisk Romance (0:58) (from Knights of The New Thunder) (1984)

reosarevok: I think once we start using aliases somewhere it will start to fall into place anyway

hawke_1: that too

that’s another good reason anyway not to use a language name but a script instead

But not that much

is that norwegian or english?

hint Klassisk is deifnitly norwegian, but "romance" is english

romanse would be the norwegian word

BrianFreud: Usually those “official” sources are more correct, though, no?

no it isn't a typo

yes, but they often don't agree

CatCat: yeah, it doesn't work for everything :)

BrianFreud: Gets back to needing a better data quality system, I guess.

what is beetle and beat in spanish?

things that work for a decent amount of stuff are much better than nothing though

BrianFreud: http://tickets.musicbrainz.org/browse/MBS-4059 is related as well

luks is right when he says we center too much on edgeish cases

well, each depends on what someone writes on a form, plus each standardizes differently - and sometimes people change titles.

a norwegian equvalent woudl be idk ,"beatillene"?

beatene

wich makes no sence

ther is a point with word and phonetic puns

it's definitly a english word

brb food refill

heh, yeah

SESAC drops all punctuation, ASCAP tends to drop "The"/"A" and spell out or drop punctuation ("#" -> "NO"), copyrighted titles can/do get changed by the filers, works get originally recorded in studio under temporary titles, etc. :)

voiceinsideyou: thanks. maybe I should start a forum thread just to brainstorm all the existing credits people have seen.

CallerNo6, wasn't there a wiki page for that a while back?

the SubOptimal Credits page, plus those annotation notes?

BrianFreud: I think he means more the variations on it

Rather than every single credit

reosarevok, I think that's because non-edge cases don't tend to need more than we already handle

BrianFreud: maybe, but they're still in their majority not in MB :p

lol, also quite true

luks' point is "let's make MB as complete as we can with non-edge stuff before getting stuck in it"

(not necessarily "let's forget it exists completely")

well, again, I thought that was the point of SOC annotation notes

Yeah, that it is

:)

But what I mean is "if MB just isn't able to do X at all, skip stuff that needs X for now and go on"

(which is basically the SOC spirit, but more in general)

I mean, there's enough work for years in Mozart + Beethoven + JSB to last someone years

reosarevok, while I am sometimes tempted to agree, I also have to ask the reverse question...

(well, a month or so, if it is BrianFreud)

s/for years//

when is MB considered "done" enough that you do then go back and try to address some of that stuff?

there's SO much not in MB that we could likely work for decades and not get to that point.

I would look at it in other way

"When is the edge case common enough that it stops being a real edge case?"

lol, even if we had just stuck to "artist/release", and never gone to label, etc.

(that's how you can easily differentiate "worth taking a look at already" from "meh, whatever")

That's also the spirit of the instrument list, for example

"Ok, this thing seems to be needed often enough for us to go and add it"

there is a flip side consideration, though; beyond the basic artist/release entry, people have different interests. What some might be interested in focusing on may bore others; the more bits of data you can handle, the broader the scope of things that might catch someone's interest, and thus increase the editor pool.

Of course, the number of uses there can be low, because it's not that hard to add it

Certainly

:)

If someone comes up with a good solution for an edge case, it's likely to get in

As long as it doesn't disrupt anything else

theoretically, at least :P

theoretically at least :p

(speaking of which, I should look at that conductor rel)

hehe

All this reminds me: luks, did you look any longer at the idea of importing releases?

it's still in the phase of contemplating, not actually working on it :)

hello

but all current or future lukz_bot edits are aimed at that

nikki is now in the working on it phase

(admittedly for a very specific use case)

hi rochusw!

reosarevok: as with everything, I want to automatically import the most useful releases first :)

not a single catalog that very few people will use

which makes it kind of hard :)

Yeah

Although I'd argue "a single catalog" beats nothing

nikki: http://userscripts.org/topics/96465 ← I can make the patch if you like, would you like ?

:)

(while you find how to do the most useful ones)

i have found useful track titles for http://musicbrainz.org/release/3c05f23c-8aff-41... (useful means, not "Sequence 123") but they are not the originl titles ( which ARE sequence...) - what should i do? create a pseudo release?

MBChatLogger, off

is not logging

is logging

I fear mistakes, many mistakes are in discogs imo and many in wikipedia … they are often copy pasted between each other, making cross checking inneffective

that's true of manually entered stuff too though

less mistakes when it comes from manual info from CD imo

if it's wrong everywhere, at least we won't be any worse :)

BrianFreud: ah you mean manual from crap sites ?

and it's easier to fix a typo then add completely new release

then yes :

luks: I input new release faster than fixing wrong release as far as I’m concerned

:p

But yeah