wolftune wrote:The only valid complete source would be entire session files. Anything less is just partial.
But like I said, what if there isn't a session file? What if I dump a GPL audio track to analog and record & mix old-school (I still keep my fostex R8 around, too many good memories there)? I could dump the analog tracks back to digital, but you wouldn't have my fader moves, eq settings, outboard FX settings, etc.
What if I bounce tracks or record with FX, and there's no way to retrieve the original takes discretely? Is it enough to release the individual tracks as they were at mixdown, or am I required to make a digital backup of each track before doing any destructive editing?
From GPL: "The "source code" for a work means the preferred form of the work for making modifications to it."
Well, if you're the person who recorded a song, your preferred form for modifying your recording is your precise session files with all the separate tracks and effect racks and everything. So, that's what full source would be when sharing the source with others.
Of course, sharing *some* source files is nicer than nothing…
Yes, it's nicer, but if we're talking about legally obligating ourselves, it's not enough; especially since the requirements would apply to anyone distributing the work (does this apply to radio stations? Music stores? Would Pandora or Jamendo be obligated to provide sources?).
If I'm obligated to provide the "preferred form of the work", who determines what's preferred? With code it's pretty straightforward, because code is basically created one way: You author a bunch of text files that are fed to a compiler / interpreter / build tool and it generates a binary.
Audio isn't so straightforward. There are a lot of ways to compile a piece of audio, and they aren't all non-destructive, automated, or easily repeatable.
My point isn't to be anti-GPL here, or to take you personally to task for wanting to release stuff GPL, but I guess I'm just thinking this through out-loud because it seems like copylefting audio (or using said audio) may ultimately be a self-defeating endeavor, if the requirements for users and distributors are too onerous or nebulous.