#musicbrainz

Harzilein, if it's not there, I guess it's just not in MB

or, perhaps more simply: there's more information that "matters" for classical, which means more decisions to make, which means more decisions to screw up :P

My point (to the extent that I have one) is that I'm all for a rich, complex CSG /if/ there's a simple way for non-classical editors to enter a disc.

yeah, I agree

reosarevok: so, how would i go about adding it? i have a badly scratched cd and lost the booklet

Harzilein, discogs is generally trusted for electronic stuff

really I wish we'd move more away from a unified "artist credit" field, or define it VERY carefully as "exactly what's on the case, and you'd damn well better add ARs to clarify"

reosarevok: but it can't just be copied to mb, right?

Harzilein, I'd go with that, maybe see if you can get the tracklist confirmed from musik-sammler or whatever

Harzilein, also, see back cover :p http://www.discogs.com/viewimages?release=265000

ianmcorvidae, while I agree, we'd need a much better UI for relationships for that

they're the same problem

we'd have a much better UI for relationships if we weren't writing UI for artist credits :)

once again, though, RDF bias -- RDF is, in some ways and translated to MB terms, nothing but ARs ;)

ianmcorvidae: Isn’t ocharles working on an AR editor?

(more powerful than that, but :P)

ianmcorvidae: Also, we still need artist credits IMO. :-p

hawke_: yeah, and if I were to bet on when it's done the time would be measured in years

ianmcorvidae, ouch :p

ianmcorvidae: And you think creating the AC editor has caused that?

I think that a focus on the special reified special-cases like ACs rather than focusing on a general case (more like ARs) has caused that

Pedant :p

sorry :P

special cases, as used everywhere. :-p

ianmcorvidae wins this round of stump the translator!

except the places they don't work :P

like all of classical

:)

hawke_, technically, a "credited to" relationship would do the same :p

and yes, that

reosarevok: Except we can’t relate to tracks

meh

yes, that's another problem -- everything should be relatable

but anyway

this is a pipedream on my part; MB decided many MANY years ago against going this route

"technically, a 'credited to' relationship would be able to do the same"

I just think we'd be better off if it hadn't :)

(it would just need to be designed for it)

Anyway, this is way off track from the problem of CSG being followed because it’s The Rules and not because it makes sense

It doesn't because the design centers on artist credits

hawke_, CSG is followed because in all its rulesy complicatedness, it gives a relatively clean set of data

Except in the works page :p

reosarevok: If you love having performers in the title, I guess

And that only sometimes

Note *relatively*

:p

it's consistent; that's what CSG is there for :P

relative to what, though? :-p

it's not fantastic because our data model is stupid for classical

as I was saying :P

there's no csg for recordings, so you can't blame the csg

relative to "put whatever you feel like"

CallerNo6: I can blame the CSG for stupid crap like 'put the performer in the title'

hawke_, that's far from being the center of CSG though

The center of it is "put the composer in the artist"

and 'don't allow the performer to be used in an AC ever'

Which admittedly you also oppose

But still :p

I don’t mind the title normalization

That bit is reasonably sensible

but the bit about how the composer, and only the composer, matters…

my argument, incidentally, has always been for "mark the performer as a performer, mark the composer as a composer, and then people can use WHAT THEY WANT"

ianmcorvidae: People want different things

hawke_, that's why they can pick their favourite! :p

yes, that's why you mark things as what they are and let people choose what they want :P

rather than fabricating _anything_ -- which, like it or not, any AC for anything classical is fabricated

CSG is there because we require an AC and so we have to fabricate something -- so it may as well be consistent

(again, practicality - helps cleanup)

ianmcorvidae: Then why put performers in the titles?

That’s not necessary.

hawke_, agreed

that's a pre-NGS normalization

(feat. xxx) was a temporary, ugly fix that predates ARs. It made sense at the time.

That can be dropped with better UI

but yes, I agree that that particular choice is stupid

ianmcorvidae: That was a pre-AR normalization

I'm saying moving to ACs will just cause different problems; the problem is a lower-level thing than EITHER of those choices

and it hasn’t been fixed since

and it won’t be fixed

well, great, so it needs fixing/changing

because CSG is The Rules

hawke_, it can be fixed

isn't that what the CSG rewrite is about? :P

It won't as long as you can't easily find the disambiguation info with a simple search

ianmcorvidae: Some people like the CSG just the way it is and are afraid of having to change things to match a new CSG

reosarevok: You can’t find shit now with a simple search. :-p

Not true

hawke_: that problem applies to any solution to this problem, including yours

There

the only solutions that don't run into that problem involve time-travel :P

reosarevok: http://beta.musicbrainz.org/search?query=holst+...

ianmcorvidae: I agree, but this is a problem for replacing CSG no matter what

hawke_, http://beta.musicbrainz.org/search?query=planet...

yes, including whatever you'd replace it with

I'm saying bringing that up doesn't matter at all :P

The idea of “we have so much data in the old format” will prevent any change to something that makes sense

With "simple" I don't mean "completely basic"

:p

hawke_, I thought the same

Then we got the feat. guideline passed

hawke_: you're right, it's just not worth paying attention to because a.) it's going to be a problem for anyone/any solution and b.) the only way to solve it is, as mentioned, time travel :P

ianmcorvidae, I'd only partially agree

reosarevok: Congratulations, your search doesn’t find the one I was looking for at all. :-p

hawke_, that means someone didn't follow the rules

(ding, that's why we have the rules :P)

reosarevok: It does?

hawke_, in most cases, it does

Incidentally, I was looking for http://beta.musicbrainz.org/release/15132986-3f...

Which looks to me like it follows the rules

lol

hawke_, it means I'm an idiot

:p

*forgot even

reosarevok: &advanced=1 is never right, IMO, for a simple search. :-p

this may qualify as the release I have the most trouble, ever, believing that it really does exist :P http://musicbrainz.org/release/6b1a4bb1-98fa-4b...

reosarevok> With "simple" I don't mean "completely basic"

<reosarevok> I mean "easy to do if you know what you want"

reosarevok: Which it’s not: 1. make a basic search, 2. scroll down, 3. check the checkbox, 4. search again

Not easy

there's too much data that's important for classical for non-advanced search to ever work -- it's like how we don't have a simple search for edits

Exactly

yes, we need an advanced search checkbox on the main search box, or to disable simple search entirely because it's unuseful

Exactly too :p

what is RDF and "reified"`

"redone"?

CatCat, http://en.wikipedia.org/wiki/Resource_Descripti...

and by reified I basically was saying "special" again :P

also to some degree "fabricated"

BrianFreud, who is the second president?

CatCat: reified means “made more concrete” (in this sense, I believe the computer science sense of “building a data model out of some ideas”)

hawke_, what we need is a nice way to search by relationships

ianmcorvidae: I call bs on “non-advanced search can never work’

reosarevok, no idea

hawke_, can it do something the advanced one can't?

reosarevok: It should be doing a more general search

I just stumbled on to it, looks like it was added in 2006 from freedb. I only added the ASIN just now when I went to see if it really actually was a real release :P

I mean, don't get me wrong. Compared with, say, allmusic's our basic search is amazing

compared with allmusic, myspace's search is amazing

But I can't see it giving anything the advanced one can't

reosarevok: But the fact that the first two results in my search don’t contain search terms I entered, while others do, is just stupid