Francisco, thanks for taking the time to report this. I downloaded your USD file and I am unable to open it in Create or usdcat. Both simply crash when I try. I’m going to dig into usdcat and see if I can get more verbose info for you.
I wonder if the crash corrupted the file and it’s just invalid now.
Based on my report below @aluk says it might be something like this:
The last time I saw a case like this, the content indices/offsets inside the crate file were corrupted to the point where Crate could not reason about the file structure (not necessarily the stage hierarchy, but how to navigate the various Crate sections denoting metadata keys, attribute values, etc).
More info:
Create’s log says this then it crashes:
[Info] [omni.usd] OmniMaterial_TreeTest.usd opened successfully in 0.01 seconds
usdchecker reports some unresolvable dependencies (http materials and textures from Create’s material browser), but nothing else
usdedit “fails to open”
usddumpcrate is able to give me a bunch of info about tokens, strings, paths, specs, etc…
Usd crate software version 0.9.0
@OmniMaterial_TreeTest.usd@ file version 0.7.0
172 specs, 172 paths, 140 tokens, 25 strings, 129 fields, 91 field sets
Structural Sections:
TOKENS 1495 bytes at offset 0x6250
STRINGS 108 bytes at offset 0x6827
FIELDS 817 bytes at offset 0x6893
FIELDSETS 377 bytes at offset 0x6BC4
PATHS 449 bytes at offset 0x6D3D
SPECS 225 bytes at offset 0x6EFE
Based on the stage hierarchy, this doesn’t look like more than experimentation (as opposed to serious customer work), but if you’re interested in recovering this stage, maybe post to usd-interest or a Github issue on Pixar’s repo to see if Pixar has any internal utilities or insight on how to potentially recover the file.
Yes, we are still testing the applications workflow with primitives and any other object that is in the create library, is not an important file, and that is what is most disturbing, it is a very simple scene, but I hope we can find what made the file corrupt in the near future.