#musicbrainz

it'd be awesome to enter those and then have MB complete the rest in the adequate/chosen locale

what's with the edits where you still have a choice to vote and it says already expired right above :s

jesus2099: nikki and I already have a private script for those links, that's why we've been using a particular format for the annotations.

I wonder what I can do to get RFV-348 passed.

warp: What change would make that acceptable to you, without having to change the schema to allow tracks without artist credits?

hawke_1: I don't remember the exact contents of 348, do you have a link?

hawke_1: but in general, a style guideline shouldn't prevent any database field from being used for it's intended purpose.

related to hawke_1 's question, if it's not a [style] question, but more a question of data modeling (as mentioned in the RFV thread) then where does /that/ get discussed?

warp: I think the problem is that the database field’s intended purpose is not so clear-cut

in addition, the old CSG specified using the composer all the time…this is at least a step in the right direction IMO.

hawke_1: yes, I'm starting to see that as well. but my attempts at getting any kind of input on that topic on the mailinglist haven't produced any useful definitions.

hawke_1: I'd suggest getting RFV-348 passed without the track artist bits as a first step.

a funny CatCat tag

hmm

warp: Hmm, it seems strange/wrong to just not specify the track artists at all.

Leaves people wondering “wtf do I put there”

hawke_: i like that sometimes

in fairness, that's no worse than the current situation :P

hawke_1: well, you can refer to the normal style guideline covering that.

warp: Do the normal style guidelines say anything about that?

hawke_1: anyway, I think it seems more important at this point to define what artist credits are.

merge advanced relationship with artist credits and get proper fields for roles

heh, thats what i thoguht At was all those years ago

and then AC

it might still happen

warp: I think we’d end up with a definition that requires a schema change though

in 3 more years

CatCat: +1.

:D

warp: So we’re effectively stuck at “can’t use the track artist field as well as we might like” but “can’t define an interim use of it” :-(

hawke_1: in most cases there is no problem, it's just edge cases like classical where people are confused, so all of it isn't terribly important ;)

hawke_1: any opinion on this? http://lists.musicbrainz.org/pipermail/musicbra...

warp: …except that people use the lack of a definition for recording ACs to say that it must be the composer.

warp: Which is a problem (at least for me)

warp: On that specific release, I agree with what others have said: the “by COIL” is definitely a track artist

the others are less clear

what others have said? I don't see any responses. :S

If a new thread were started to talk about ACs, where would that be? [style]? [dev]?

CallerNo6: talk about ACs how?

also probably style, dev is usually pretty dead :/

sorry, talk about the definition of ACs

(since elsewhere in that thread warp suggests that it might not be a "style" issue per se)

probably style, not that I'd expect that to result in much of a conclusion if any

warp: Heh, whoops. I guess others was you. :-D

honestly, the place to discuss it might be here

I don't think we have a real place for discussing semi-style semi-dev issues except IRC

hawke_1: so do you agree the other tracks and the release do not have artist credits?

[unknown]?

warp: Yes, though I wouldn’t object to considering the “Music:” or “Music/Arrangement” to be artist credits. (as well as ARs)

hawke_1: I guess my problem is that doing so confuses the concept, and I think it would be clearer if we could just have [none] if there is none.

(and then picard can substitute the composer(s) or performer(s) based on user preference when tagging)

warp: I don’t think we have a system in place to record or describe that user preference

warp, that's /if/ the concept is in line with your idea of ACs, i.e. a factual field based on tabular data found in a particular part of a release

warp: There’s something to be said for treating it as an answer to the question “what artist would you say this recording is by”

saying it’s by [nothing] is not helpful to anyone

of course, that doesn't really work for tracks (since tracks are a weird amalgam of release, recording, and work)

Yes

but I think for recordings, I agree

ianmcorvidae: The problem with using it for recordings is that there are often many answers to that question. :-)

CallerNo6: certainly. but I haven't heard any other views, everyone seems to infer that some other view exist, but no one is willing to state it.

warp: the other view I've heard is "as close to whatever you use for the track as possible"

I really hate the idea of recordings being credited to someone who had nothing to do with the actual process of creating the recording

(which seems to me like people overly used to classic MB)

warp: I think I've talked about my view in other threads (or maybe here in IRC). I'll respond in full on [style].

CallerNo6: thanks.

ianmcorvidae: "as close to whatever you use for the track" doesn't make any sense when trying to determine the artist credit for a track.

warp: I was talking about recording ACs

as far as I know there's no opinion but "on the case, when applicable, otherwise who the hell knows", as Track ACs go

ianmcorvidae: yeah, so I think the "who the hell knows" part should just be [none] to make things clear and predictable.

I think the AC should be optional, yes

I'm ok with using the release artist on tracks, which would result in track ACs and release ACs only being [none] when the release artist is [none].

I’ve posted my response to -style.

warp: What about Various Artists?

I'm also ok with not using [none], but then my idea of what an artist credit is off, and then I need to be told what an artist credit is :)

hawke_1: I expect most VA releases are not in general credited with "Various Artists", so it would be more accurate to use [none] there.

warp: That’s true of pretty much all VA releases in existence, yes.

But [none] doesn’t help anyone, any more than VA does.

They’re both just placeholders

Yeah

I think a question is: what good is it to tell people that the release doesn't have an straightforward credit?

(if any)

And I think it’s silly to use [none] when you have only one possible artist

(as on most of the tracks of your example release)

but I think it would be inaccurate to credit this particular soundtrack release to "Various Artists", most of it is likely composed and performed by a small group of people.

warp: I’d certainly credit it to COIL, yes.

hawke_1: oh, I was in the process of starting a new thread. to separate the AC general discussion from the CSG question

so then the concept of a release artist credit is becoming more fuzzy again.

oh well

warp: it will always be fuzzy

warp: Sure, because the “use the artist that has the majority of the track credits” is helpful.

(at least when there is one artist that has a clear majority)

hawke_1: I'm not opposed to the substition of [none] with COIL in this case, but I think it would help if that substitation is codified in our style guidelines.

The answer b) would be "use artist credits as a way of navigating, and use whatever is better for the user to use the data for both his own use and database navigation"

Since it's ill-defined anyway, "if there's no guideline use what you'd expect to see this filed under" seems to make sense to me

reosarevok: In that case this release would probably be under [soundtrack]

Wait, anyone actually uses that for sorting?

Curious

reosarevok: I see record stores with soundtrack sections, sure.

reosarevok: that sounds far too subjective.

warp: because artist fields are pretty much subjective

What they decided to credit is also subjective

also “filed under” is kind of an anachronism with searchable databases.

The factual stuff is in the rels

reosarevok: no they aren't

There's little more subjective than which name is bigger in a cover, or which one is relegated to the notes

I mean, that's useful info, but isn't that why we're going to allow people to see it from the actual cover?

CatCat: based on what?

Dunno, if I have a white label vinyl and I know what's inside, I wouldn't want it filed under [none], I would want it filed under whatever is inside, for example

Even if there's not a single letter in the actual release

reosarevok: I'm not saying using [none] is particularly useful. I'm only saying it would be accurate.

but "artist" should be a field with a name, eg "composer" "performer"

"guitarist"

etc

Except that as you're seeing yourself, it wouldn't, because the limit of what is an artist credit and what is a note is fuzzy

reosarevok: I'm not seeing that yet :)

its the thing what is a we want, what it is credited or what it actually *is*

So one release would have some stuff as credits, and other would have it only as rels, depending on where the adder though that fuzzy limit was

warp: https://frob.nl/mb/2012/KECH-1190/booklet01.png

Is track 7 an artist credit?

If it is, who's the artist? "feat. Nahki"?

(there's no main artist there)

Is it COIL feat. NAHKI? (then you're making the credit up)

reosarevok: I would use just "Nahki" as the artist credit, yes.

warp: well on 8 it says "Music : COIL" so COIL is the artist, on 10 it says "music/arrangement : 原田智弘" so.. 原田智弘 is the artist?

i find the "by YUAN or by COIL" to be way more ambigus

what exactly "by"?

warp: except that other person would use COIL feat. NAHKI because "feat" assumes there's a second artist

reosarevok: based on what? COIL isn't even the release artist, where did they get COIL from?

i'd use "featuring", cause japanese

In that it's the only other artist in that track

(credited)

i see "Music : COIL"

So if an artist is featuring NAHKI it must be them

(or him, or whatever)

coil is a good word btw

I like it

reosarevok: a person doing that would presumably in general stick stuff which should go in relationships (also) in the artist credits if that person couldn't find anything better.

warp: HEY!

warp: I think that’s the point. ;-)

Not necessarily. It's just the word featuring that semantically asks for a second artist

(although that's also possible)

if it isn't then it's i nthe trackname then

Also, who is this credited to? http://imgur.com/CrB5v

reosarevok: which brings us back to what I'm trying to say here, that if we are going to substitute a missing artist credits with stuff from liner notes / relationships, we should have some guidelines for it

"nuff REspect featuring "NAHKI""