Opinions: Collabs With Connectivity As Single Units

I've been holed up in my Fortress Of Solitude for the last week trying to hammer out once and for all the Linking / Family Tree situation when dealing with collabs with connectivity (CC's). After a few different implementations I was looking at the possibility of "Neil Young & Crazy Horse" (otherwise known as a CC) as a distinct entity in family trees. If a tree were to theoretically look like:

Neil Young -> Neil Young & Crazy Horse -> Crazy Horse -> The Rockets

it seems redundant for both instances of Neil Young to point to Neil Young and both instances of Crazy Horse to point to Crazy Horse. That of course leads to the idea, that people here have preposed on a number of occasions, that Neil Young & Crazy Horse should be it's own entity and if you were to click on it then you would get all the releases by that CC. In other words, make it a true band as we do for Tom Petty & The Heartbreakers (not a CC).

This topic has been discussed before, on many occasions, but instead of trying to go back through those old topics I'd thought I'd start a new one and just get a head count of everyone who thinks CC should be replaced by "true" bands, ie Neil Young And Crazy Horse is a distinct band from Neil Young, which is distinct from Crazy Horse.

I'm still not excited about situations like "Brian Eno - David Byrne" being its own band, although that would be a by-product of the change in approach, but then again I'm not too excited about "Neil Young & Crazy Horse" not being clickable in a family tree.

So, just throwing it out there, who's for CC's being their own band and who is for the current system of keeping both ends distinct from one another?


Kevin

CC's should be their own band. Always been my vote.

I vote for it too.

As long as it's restricted to the family tree, I vote for it.

I agree, CCs should be their own band.

I'm not sure I wouldn't go even further and make all collabs CCs, given that the "members" of such collabs often work together more closely than some of the "real bands" we have (prog metal supergroups, I'm looking at you!). The complexity of the collab rules has always boggled me.

I'm definitely for CC's in the family tree, because the family tree's purpose is to expose connectivity, and CC's are a part of that.

I'm less sure that I'm for CC's being their own band, but I think I'd land on the side of them being bands. I wouldn't be unhappy if it excluded CC's made up entirely of solo artists, though. So I guess I'm ambivalent on this.

It seems that everyone is on the side of collapsing CC's into single units so I'm going to work towards making the Big Collapse happen.

Python:
What do you mean by just restricting it to only family trees? Should Neil Young and Crazy Horse still be distinct entities on the album "Broken Arrow" for example?
[bandtoband.com] http://bandtoband.com/index.php?Page=Search&AlbumId=12794

AJ:
I'm sort of with you on the solo artist + solo artist but for the sake of argueent what about the case of Richard & Linda Thompson?
[bandtoband.com] http://bandtoband.com/index.php?Page=Search&AlbumId=13632
While solo + solo a good case could be made that it should be its own band.

I will say that part of the allure is being able to reduce the Rule complexity for collaborations down to:
1) A collaboration exists only If all parties involved released material independently of on another before the collaboration and in that case there never is connectivity.


Kevin

That would be a substantial change. At the moment only one part of a collab can have released independent material - you are now as I understand it proposing to reverse that and the presence of 1 previously unrecorded artist would turn this entity into a discrete recording unit.
Did I understand that correctly?

Re: Richard & Linda Thompson

Like I said, I'm not married to this concept (ha!).

I agree with the feeling that "Brian Eno - David Byrne" isn't really a band, and the collapse would make it one. But I don't feel strongly enough about it to make a big deal over it.

The answer in my mind, without respect to solo/solo vs solo/band vs band/band, is to be more holistic about it, and ask questions like: Did they collaborate multiple times? (If so, probably a band.) Were all of them "new" entities, or was one of them established already? (If one was established before the collab, it's not a band.) And so on. Sadly, this does not reduce to a nice rule very well.

Here's a couple of collapsed CC's:
[bandtoband.com] http://bandtoband.com/index.php?Page=Search&BandId=9120
[bandtoband.com] http://bandtoband.com/index.php?Page=Search&BandId=6344

Just to let everyone know what they're in for...


Kevin

Yikes, I'm not so sure I like that. Maybe I'm just too used to seeing collabs in the old style. More cautiously, I'd like to see the change only in the family tree (as ajw originally suggested).

Just to be clear, here's what I had envisioned for the family trees when we started down this path. Neil Young's tree would look like this:

Neil Young -> Neil Young & Crazy Horse -> The Rockets
......................................................-> Crazy Horse

And Crazy Horse would look like this:

Crazy Horse -> Neil Young & Crazy Horse -> Crosby, Stills, Nash & Young
........................................................-> Neil Young

Perhaps "Neil Young" and "Crazy Horse" could be individually clickable in the CC, as they are on album pages. That would allow us to keep our old collab rules and still have family trees displayed sensibly (i.e., a branch between two nodes indicates shared membership between named entities).

I like the examples shown and continue to think is is a good idea.

The VU & Nico thread in Complete Shambles suggests Kevin is already changing all CCs to bands. So is this still open to discussion or is it already a fait accompli?

I've done a handful so far, and although I have a program written to crunch them all I've been taking my time to make sure nothing goes too crazy. I'd agree that I'm not to excited about
Art Blakey's Jazz Messengers With Thelonious Monk
being a single band when it seems clear that "With" is used to distinguish the two parts. Then again in order to collapse "NY & CH" or "Bob Marley & The Wailers" (which has been done for at least one album)
[bandtoband.com] http://bandtoband.com/index.php?Page=Search&AlbumId=7960
it's the price that you have to pay. The only way to thwart the Thelonious Monk situation is to find a recording credited to Monk before that CC.

Another argument for the Collapse is having "Neil Young & Crazy Horse" as an entity to view distinct from both Neil Young and the band Crazy Horse. That way you can see what Neil Young has done by himself, what Crazy Horse did by themselves, and what they did together. I'm sure there is some insane complexity I could introduce into the system that could force a distinction between CC's but reducing complexity is always a nice thing.

Let me know if there are any more reservations out there before I continue on...


Kevin

I guess I can live with the new post-Collapse arrangement, if nobody else objects to it. (But it makes my head hurt thinking what the new collab rules will be like and how they'll affect submissions in the queue...)

While initially I wasn't a big fan of the examples you posted, after thinking about it for awhile I think collapsing all these makes the most sense and is the simplest and most consistent way to treat all the data. (Which is probably good since I'm late enough to this discussion that you've already done it all anyway.)