[x] Collabs

Is the styling of a collab set on its first release? For example if Bob Smith & Bill Jones release first as
'Bob Smith & Bill Jones' do we apply that as primary for other releases by 'Bill Jones & Bob Smith' or 'Bob Smith / Bill Jones'

Taking that further if they then release as Jones/Smith - what do we do? That's not a collab by strict rules but (IMHO) by any sensible interpretation it's just another styling of the original name.

i think this was previously discussed here:
[www.bandtoband.com] http://www.bandtoband.com/forums/viewtopic.php?t=5182

in your example i guess 'Bob Smith & Bill Jones' becomes the default band name. if the next release was 'Bill Jones & Bob Smith' it would be:
Bob Smith & Bill Jones||Bill Jones & Bob Smith.
if however, the next release was 'Bob Smith And Bill Jones' this might become the default band name as, in my experience, we seem to default to 'And' rather than '&'. not sure of the exact hierarchy of these connectors but it seems to be, from weakest to strongest:
With < & < And < / < [cdot]
correct me if i'm wrong...

not sure about Jones/Smith. i would tend to agree with your reading, which i think would mean, in the case of collabs, Kazuyuki K Null = KK Null and Polly Jean Harvey is the same as PJ Harvey.
not sure if this is really that consistent with how things are usually done on the site (we would treat Kazuyuki K Null and KK Null as two different artists for solo purposes) but i agree with the sentiment
"Once a group of people collaborate as a band (without prior solo), they're a band, despite the relative name ordering". i would extend that also to "relative name crediting" within reasonable bounds.

I agree with most of shakinghell's points - I was going to post a link to that previous discussion too.

Off the back of that conservation I've aliased a couple of collab releases as suggested e.g. this one with aliases:
[www.bandtoband.com] http://www.bandtoband.com/index.php?PageField=Wiki&amp;ActionField=var&amp;Album...
after the original ordering was established here:
[www.bandtoband.com] http://www.bandtoband.com/index.php?PageField=Wiki&amp;ActionField=var&amp;Album...

However... As for dropping surnames or switching from first names to initials etc, while I agree it seems logical to allow a certan degree of aliasing rather than creating a lot of new bands with idenitcal membership due to some fairly minor crediting changes, I think we have to be consistent.

For me, if we treat Kazuyuki K Null as a separate solo artist to KK Null then we should treat them as separate artists in the context of collaborations as well.

Don't disagree that those are the rules now but won't we just look odd to anyone coming along. Our rules will appear to have landed us in a very silly situation indeed. So, you collaborate for 30 years as 'Bill Smith & Bob Jones' - but both Bill and Bob had one obscure release before the first collab. No linkage we say. Then you release one song as Smith & Jones and we leap upon it as linked. That is another argument moving me towards a position where I'd happily get rid of collabs all together if it were ever voted on.

I don't disagree - our current rules can create a major difference between, from your example, 'Bill Smith & Bob Jones' and 'Smith & Jones' when in reality the difference is minimal, and I'd bet if either of these fictional characters had a discography on their websites they'd list both releases as being by the same 'band'.

In exactly the same way we would currently differentiate between an album by 'Bill Smith - Bob Jones' (not a band if both have prior solo work) and 'Bill Smith - Bob Jones Duo' (a band no matter what) simply because they bothered to add the word 'duo' to the album cover. Logically this doesn't seem quite right.

When you suggest getting rid of collaborations entirely, I assume you mean just removing the 'one collaborator must have no prior independent releases' rule so that even well established artists working together would create a new 'band' (and 'Bill Smith & Bob Jones' would be a band regardless of whether either had prior solo work. That approach is starting to strike me as increasingly sensible and I think it's worth considering.

a little off topic but i thought of a slightly different approach to collabs that may allow some greater connectivity. One concern is that it is hard to actually track if someone has had any solo material out previously, especially when it comes to obscure multi-band compilations and, especially in the genre of hip-hop, semi-official mixtapes.
well, we don't put as much emphasis on the crediting style on multi-band compilations anyway, so maybe we could exclude this format when it comes to determining if an artist had previous solo material out.

Steve Albini, for example, had one 1'46-long instrumental guitar track credited to him on this compilation:
[www.bandtoband.com] http://www.bandtoband.com/band/steve-albini/guitarrorists
but i hardly think this makes him an established solo artist. nevertheless, this entry prevents the following release by Zeni Geva & Steve Albini as being counted as a band:
[www.bandtoband.com] http://www.bandtoband.com/band/steve-albini/all-right-you-little-bastards

perhaps it would be easier and more consistent to look only at albums, singles and splits/shared releases when it comes to establishing solo artist status.

i'm sure there are arguments against this - one being that, though we ignore multi-band comps in some cases, we are happy to include 'bands' such as these from those same compilations:
[www.bandtoband.com] http://www.bandtoband.com/index.php?PageField=Wiki&amp;ActionField=var&amp;Album...
also, we are happy to take advantage of those naming mishaps more common on multi-band releases to connect artists as a band. can't think of an example we have exploited, but what if a track by Peter Gabrel / Kate Bush ended up on a comp.? a new Peter Gabriel alias ('Gabrel') and connectivity?

anyway, that's an approach to think about which takes us closer to eliminating collabs altogether. (FWIW i'm still divided about that idea. something about the idea that even Eno / Cale is a real band and not a collab doesn't sit well with me, but on the other hand i'm happy to have Crosby, Stills & Nash as a band)

Regarding the beginning of this thread, I'm posting this in several places:

[bandtoband.com] http://bandtoband.com/forums/viewtopic.php?t=5182
[bandtoband.com] http://bandtoband.com/forums/viewtopic.php?t=8355
[bandtoband.com] http://bandtoband.com/forums/viewtopic.php?t=8589

Band aliasing is supposed to be limited to just trivial name changes. We've never treated order changes as trivial. As in [www.bandtoband.com] http://www.bandtoband.com/forums/viewtopic.php?t=6169 (the last post), B/A and A/B should go into the database as distinct entries. Why? Because they are distinct. A and B could have made sure to release all their albums as either A/B or B/A. But they didn't, and we should document that. Why would we try to force the two bands into one if they didn't bother to do it themselves?

---

As for shakinghell's most recent post, we recognize all other "bands" that appear on compilations, so we should continue to recognize solo appearances on compilations.