> Distance is ideally user-variable. When you build an object, sometimes you > want very dense spacing because you want a very strong object, and sometime > you dont care so you space it very widely.
I absolutely love this idea of a space filling algorithm, that would be a massive advantage… I recall STL softwares building automatic supports, but haven't heared of strategies for dealing with solids… Ran into this problem quite a few times myself ;') > > Another odd quirk is that its actually better to have the hatching lines not > be parallel. Printed objects are stronger with solid walls. So, suppose you > have inner fill lines that are space 3mm apart. what you want, is for the > lines to always end up very close together near the walls, but be further > apart in the center. For sure. This wouldn't be too difficult to pull of… just offset the slice following some gaussian fall-off, right? > >Yes, given a high level precision of the machine ( where machine_precision < > >tesselation_error ) it sure could be interesting… > Also, it would be less imposing on your hardware controllers, right? > > Definitely. as an example, I created a circle with diameter 2 inches, and > exported it to STL with tolerance of 0.001. The resulting file was 450 lines > of G01 moves. My machine's precision is probably on this same order: accuracy > 0.0005 but precision 0.001 due to backlash. of course the g02 move for the > same motion was 1 line of code. Funny… > >You did see that we paralleled your slicer, right? > No i didnt notice that, but i do now. I see you used BRepAlgo_Section, which > i suspect works must faster than my method. I ended up giving up trying to > use BRepAlgo_Section because I couldnt find a good way to assemble the > resulting compound back into something i could work with. For example, > testing a point to see if it is 'inside' or outside a face, etc is easier > when i have a first-class face instead of a compound. Its amusing to see it run on a 8-core machine ;') Also, the display.DisplayShape( some_brep, update=False) got a new update keyword, accepts iterables. If you add lots and lots of geometry, this speeds up things _dramatically_ ;') > Thank you for your help and samples below. I will definitely make you guys a > part when i get it done! Weeeeeeeeeeeeeeeeh! Cool! I'm amazed by the precision of your machine, is that is RepRap machine? Amazing project Dave! -jelle
_______________________________________________ Pythonocc-users mailing list Pythonocc-users@gna.org https://mail.gna.org/listinfo/pythonocc-users
