Profiles proposal: Can a base profile describe a non-rendering domain?

Community USD
1

Hi there! First time caller here.

Long-time USD user on the DCC/Solaris side.

I’ve read the Profiles
proposal (merged in OpenUSD-proposals:
OpenUSD-proposals/proposals/profiles at main · PixarAnimationStudios/OpenUSD-proposals · GitHub)
and the USD WG agendas that reference it, and I have one scope question.

Two things in it caught my attention. It suggests future base profiles
could be managed by bodies like AOUSD, with a limited web profile
(org.aousd.web) as the example. And the Runtime Analysis section already
contemplates consumers that aren’t rendering at all — “a runtime that is
not rendering” can skip the visualization categories entirely.

My question: is the intent that base profiles are always subsets of
scene-description USD, or could a base profile legitimately describe a
domain where nothing in the stage renders — where composition and value
resolution over layered data are the point, and the schemas are
domain-specific rather than geometric?

Asking to calibrate scope, not proposing anything yet. If the WG has
discussed this beyond what’s in the meeting agendas, pointers
appreciated.

PS… Mods: let me know if I’m posting in the right spot.

Best Joseph

2

I moved this to a more generic section since profiles themselves aren’t in the core specification yet.

I’d defer to @nporcino to respond more in depth, though you might have to wait post SIGGRAPH.

In general, the way our USD specs are set up is that yes in theory you could make a file or profile that has no renderable elements. Thats possible today as well of course with spatial audio only files etc… but its also one reason we split the core spec from any renderable domains like geometry so that it would be possible to pick and choose domains as long as the core is honoured.

3

You are reading it as intended Joseph :slight_smile: A profile could “legitimately describe a
domain where nothing in the stage renders — where composition and value
resolution over layered data are the point, and the schemas are
domain-specific rather than geometric”

4

Thank you both for your quick response. This gives me alot to think about. Enjoy Siggraph!