Hey,
I've used the Bound Body Hull -> Convex Hull option to create a stone from a random convex hull that I inflated (Really fast way to get some decent looking rocks, btw), but when I try to UV Map this shape, I get the following crash dump (See attached file) when I select Continue -> Project Normal in the AutoUV Segmenting window.
Didn't see any UVMapping cues to be used by AutoUV.
What cues did you provide for cutting ?
OK ... this look like a possible problem with an "illegal name" created by ManifoldLab.
I don't know this with 100% confidence ... but if you proceed as if true ... maybe you can
find more of a smoking gun. Renaming object also failed me. I had to rename by using
Manifoldlab rename.
Try any sort of name washing ... including export to OBJ ... look at names ... re-import. We will fix this
once we do some basic tests here.
Lists in Erlang are "FUNNY" ... it is easy to product a list of lists that sometimes can pass onwards as a string ... and sometimes not.
If you could list the exact steps you take to make a rock that fails ... I will debug it all the way in ManifoldLab and release a fix.
Also ... let me know what build/version/date of Manifoldlab (about menu) you used when having issue.
I encountered this issue myself a few weeks ago. It MIGHT be fixed in latest manifoldlab. Not sure.
Either way ... to fix ... we need simplest receipe for producing the bug.
I was using version 2015-02-21--18.18.mlab
My steps were (To the best of my memory)
1) Create a convex hull, random
2) Select verts -> Bounding Body Hull -> Convex Hull
3) Repeat steps 1 and 2 a few times.
4) Select verts -> Deform -> Inflate
5) Use turn edge on a few of the edges
6) Used MLab Rename Sequence (Probably, like you say, it was the bad naming?)
7) Combine objects -> UV Mapping
Select all faces -> Segment by -> Projection
9) Continue -> Project Normal
10) Crash
I ended up exporting as an obj and unwrapped in Ultimate Unwrap, then re-imported the mesh, and were now able to see the UV Maps again.
Nova ... check out this relevant reading ...
http://stackoverflow.com/questions/65032...-in-erlang
I'm guilty of using that io_lib function ... and it produced a deep list. Which in fact is handled fairly welll ny a number of parts of Wings3D ... so that the potential problem can haunt at a much later date. Anyways. I know how to fix it in MLAB and will do so soon. I'm just fixing the rename sequence function tonight to remove the deep list.
New release just now ... to fix the problem with the rename sequence.
I argued the other day that it could be an AutoUV fix ... but probably better to
squash the issue at it's source.