[mb-style] Band name changes (was [mb-devel] Panic! at the Disco)

[Chris B]

On 19/03/2008, Lauri Watts <krazykiwi at gmail.com> wrote:
> On Wed, Mar 19, 2008 at 4:06 PM, Bram van Dijk
>  <bram_van_dijk at hotmail.com> wrote:
>  > Well, http://www.aelius.com/njh/tmp/musicbrainz_summit8_schema.pdf
>  >  seems to indicate that we will (eventually) get something like Discogs'
>  >  ANV system.
>  >
>  >  But I agree that until then, it may be better  to keep the releases
>  >  together.
>
>
> I actually think the opposite: Until we can indicate which is by which
>  name when they're on the same page, it's better to keep them separate.
>
>  Untangling which release has which name from all those compilations
>  alone, is going to be a nightmare for some of these artists.  If we
>  kept them separate now, but AR'ed, users could still follow the trail,
>  albeit not as easily as if they were all on one page, and we aren't
>  setting up yet another pile of rather tedious work once NGS is
>  available.

it'll require work either way. if it went:
The Vandellas
{changed name to (in 1970)}
Martha & The Vandellas
{change named to (in 1975)}
Martha Reeves & The Vandellas
how do the ARs link? to each other in a chain (would seem logical in
the context of the AR alone) or all back to one primary name (and
which one?).

currently we would (all else being equal) merge to the most useful
(ie, popular) name, which has manifest benefits for both the
discography and the tagger (as previously mentioned). if we retain
that (ARing to one rather than linking in a chain), then at least the
tagger side can be accounted for by having some kind of preference in
picard to use primary names of artists when tagging. is that your
intention?

i don't buy that we should 'prepare' the data for NGS - isn't that the
reason we're having NGS in the first place? because such an 'accurate'
DB structure is unsuitable for our needs without major DB and display
changes. obviously there's going to be significant editing work across
the board once we go to NGS from our current 'amended' data anyway...