I applied this on Blender 2.72 and got a LITE version to work. It was very useful.
I applied this on Blender 2.72 and got a LITE version to work. It was very useful.
Complexity not in the code, but in the flow and allocations. It's kinda overkill to have dedicated CD allocation type to support some particular case of DM to mesh conversion. It'll be more useful i think to have function which would ensure DM does not have referenced layers. It'll be much more clear what's happening with data ownership.
Well, do not see much complexity in this, it's pretty straightforward and useful imho…
Just tested on my laptop.. i72.3GHz with nvidia 630m, ubuntu14.10 and can't reproduce the crash there. Maybe I'ts my desktop pc asking me to buy a decent gpu. If anyone else can't reproduce this, feel free to close this issue. Hope I didn't make you guys waste to much time.
This regression is related (i think) with the change to SDL2.
Yep, we should have FindLLVM.cmake
yes, yes, and not quite ready yet. Much improved, need to finalize a few details coding wise; and working on wiki to match use and operation. Thank you for all your support.
Hi, thanks for updating, I've no problem putting this into add-mesh_building objects in contrib.
I believe you already have git access.
We can ask mont29 to quick view the code & approve it's re-inclusion.
iirc columns was in old svn a few years back.
Can't redo this bug, menu navigation works fine here.
Cleanup blender code style.
Ready for next round of "how does this work" - to include replace current py script and doc. ;)
add_mesh_columns.py
Couldnt redo the crash.
Arch linux, 2.73a release (or a recent debug build), both with CUDA and CPU rendering.
First of all, it'll break someone's day, plus misses changes in install_deps.sh to advertise proper CMake settings now ;)
Closed by commit rBda0176614b37.