Hi Stefan, yes testing with your posted model there, the issue is that you're running out of memory.

The way NURBS models are structured with a system of an "underlying surface" and trimming boundaries having several components like a UV and a 3D part of each edge in the trimming boundary, there is a higher amount of overhead in each individual surface. So things are generally meant to have larger areas of a model made up of large smooth surfaces and not having a structure of something like a hundreds of thousands of little separate facet pieces like you would have in a polygon mesh model structure.

That's why it does not generally work very well to try and convert a dense polygon mesh model directly in to a faceted NURBS model.


But now that I've done a bit of testing, in this particular case it appears that it's actually the history mechanism that is actually the first thing that's running out of memory though - a record of the inputs that went into the command is attached to every output object that is generated, and in this case since you've got so many input objects going into the command all at once each history record is pretty large and then since you also have so many output objects having that large record replicated many times on each generated object causes a substantial amount of memory consumption.

In the future I want to restructure how history works so that it's recorded in one central spot as more of a list of actions rather than tagged on individual objects like it is currently. For the next beta I'll modify the Planar command so that it won't try to save history if you've got a large number of objects being processed, that should help with this particular case.

For the moment you'll have to run Planar on a somewhat smaller set of objects simultaneously in order to avoid running out of memory due to history storage.

- Michael