is there any way to get this component to work faster? I’ve tried making it multithreaded using both python and c# but it usually crashes or random patches get left out.
maybe a shorter term solution might be some kind of progress indicator to know whether it’s actually calculating or if there’s really a crash
wow 1.8 h, you can increase cutting by the following:
check cp count and see if you can simplify geom and cutter
sort your cutting curves in a way that they always just get the right surface to cut
multithreading (be sure about threadsafety)
if nothing helps: pro solutions:
map your surfaces and curves in one base plane, use surface from boundary, copy resulting trim into your 3dspace curves
or make sure your surfaces are on the geometry and create a trimmed brep manually. If you do this, you can bypass all boundary checks.
if nothings helps: pragmatic solution:
bake and use split command. you can divide your geometry in sectors saving in different 3dms, open many rhino instances and split manually in each instance. This is multithreading as well.
Can this process even be multi-threaded (aside from your multiple 3dm idea) since you would need to re-use the surface after each cut? Seems you need to do either all at once (as it is now, not multi threaded because the circles need to know about eachother) or one at a time but by reusing the surface from the previous cut.
I was primarily refering to the second cut, since the first isn‘t the main problem. depending of the amount of cutting operations, 1.8 min is acceptable. without concrete geometry its hard to judge. you can of course merge it into one splitting, or you need to sort cutter and first level splits befor cutting a second time. if you can provide a file with geom its better to judge, although I won‘t have much time to have alook at today, but maybe others can help
I don’t have any particular geometry, maybe @arch can share his. Would be interested to see some potential threading. Was just curious because I though this kind of task was against the nature of multi threading but I think you know more about that than I do
@arch If it helps at all there was a discussion about this some time ago and was found that using the split method in a loop and cutting one at a time over and over was much faster.