Can confirm.

Thx for investigating!
Could you share a minimal example script that leads to the crash?
(this will make it much easier for devs looking into this to reproduce. thx in advance!)

In T77046#954666, @Ray molenkamp (LazyDodo) wrote:

no repro with 2.83 or 2.90 even tried the hardened heap.

Given you have the crash with 2.90, when you run it from the command line, and it crashes it should tell you the location of the crash report (generally somewhere in the temp folder) please attach this file to this report.

Remove redundant set to 0.

@Julian Eisel (Severin) This might be for you, since you've originally implemented this feature.

Concerning line endings:

Thanks Nathan. Sorry, I had a mind shortcut and transformed Rhinos3D 6.0 to Rhinos 3.0. Indeed, it is Rhinos 6.0 the original file.

fixed another None Type error in importer - do not know where that None type is coming from

Right, binary format indeed fails.

changed object list to square brackets, otherwise no objects get exported

Thanks. The files you provided are in ASCII format, whereas the original one is in binary format. Can you see if you can select the export option as binary in 6? Otherwise we do not know if the bug is still there in 6.0. Thanks!

FWIW: current stable version of Rhino is v6 (SR 27 I believe).

Thanks. In case it helps, Meshlab uses the following chars as possible separators:
" \t\n\r"

Will confirm for now, apparently there is an issue with line termination, yes.
Not sure if it is against the specs, but using line in plyf [plyf is open(filepath, 'rb')] will not separate lines properly.
Tried using plyf.read().splitlines() - that separates them properly - but has issues later on.
Needs more investigation...

Fixeed keyframe chunk definitions

fixed some wrong intends

fixed non-type error in world color if no world is used

I'm unable to reproduce the bug on Linux 64 bit, build d290bdd42abb (2020-04-19) from buildbot. Running the script several times and modifying the wait time never caused a crash.

found a little error in the chunk definitions

In D7454#197060, @Campbell Barton (campbellbarton) wrote:

@Sebastian Sille (nrgsille), would you be interested to have commit access to the add-ons repository to maintain this add-on?

@Sebastian Sille (nrgsille), would you be interested to have commit access to the add-ons repository to maintain this add-on?

Well that's strange.. This is what happening to me with a Blender in factory settings :

• First, I I use ALT+LEFTMOUSE, you can see the invoke is called every times, I need to RIGHTCLICK multiple times to cancelled the modal.
• Then I use S+LEFTMOUSE multiple times but you can see that the invoke is called only one time.

use the combination "S"+LEFTMOUSE

I don't have the time to write docs for this. Maybe some day ;-)

Not sure about abc exporter, but AFAIK feature to access render depsgraph(which is also separate) would be nice to have, but I am not aware of any concrete plans.

In T74339#955112, @Richard Antalik (ISS) wrote:

I can not reproduce any crash here as well (using 2.90 build). Can not check 542ff416e274 to see if this has been resolved. Can you check again with latest 2.90 build?

The fact, that exported mesh is static is likely correct, since alembic exporter uses own dependency graph.

See https://bugs.python.org/issue18091 as well.

The simple script has already been provided (see history of this conversation : May 25, 3 PM)
And (May 28 6 PM) -> you can download the zip with the code inside...

This report requires far too many steps to redo, please include a sample blend file that includes a script that crashes on execution.

fixed spotlight triangulation