I don't know the default configuration: did you turn on all the OpenSCAD features that are related to faster rendering? A lot of times the new features in the dev versions are off by default. It's all the fast-csg things. For me the 3 blade case rendersin 2022.12.07 in 0.35s. It took 4:14 in the stable version of OpenSCAD. On Sat, Feb 18, 2023 at 9:20 PM Father Horton <fatherhorton@gmail.com> wrote: > I downloaded and ran 2023.2.15, ran each one twice, and flushed the cache > between runs. > > With caps: > 1 blade: 0:21 both times > 2 blades: 1:14 both times > 3 blades: 3:11/3:18 > > With caps: > 1 blade: 1.4 sec/.4 sec > 2 blades: 1:07 both times > 3 blades: 3:31/3:35 > So the dev release does render faster, and I have no idea what effect the > caps are having. The exponential time increase makes sense if OpenSCAD is > having to check for collisions of more objects or something like that. > > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

I didn’t turn them on. I will try again later. On Sat, Feb 18, 2023 at 9:06 PM Adrian Mariano <avm4@cornell.edu> wrote: > I don't know the default configuration: did you turn on all the OpenSCAD > features that are related to faster rendering? A lot of times the new > features in the dev versions are off by default. It's all the fast-csg > things. > > For me the 3 blade case rendersin 2022.12.07 in 0.35s. It took 4:14 in > the stable version of OpenSCAD. > > > > On Sat, Feb 18, 2023 at 9:20 PM Father Horton <fatherhorton@gmail.com> > wrote: > >> I downloaded and ran 2023.2.15, ran each one twice, and flushed the cache >> between runs. >> >> With caps: >> 1 blade: 0:21 both times >> 2 blades: 1:14 both times >> 3 blades: 3:11/3:18 >> >> With caps: >> 1 blade: 1.4 sec/.4 sec >> 2 blades: 1:07 both times >> 3 blades: 3:31/3:35 >> So the dev release does render faster, and I have no idea what effect the >> caps are having. The exponential time increase makes sense if OpenSCAD is >> having to check for collisions of more objects or something like that. >> >> _______________________________________________ >> OpenSCAD mailing list >> To unsubscribe send an email to discuss-leave@lists.openscad.org >> > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

On Sat, 2023-02-18 at 22:05 -0500, Adrian Mariano wrote: > I don't know the default configuration:  did you turn on all the > OpenSCAD features that are related to faster rendering?  A lot of > times the new features in the dev versions are off by default.  It's > all the fast-csg things.  > > For me the 3 blade case rendersin 2022.12.07 in 0.35s.  It took 4:14 > in the stable version of OpenSCAD. was that with or without the caps. I used dev snapshot appimage version 2023.02.15.ai13654 (git 9e8fa3dc5) I am wondering if the strange speed results are because of the caps having mesh problems. My results, first run without caps, then with: Caches Flushed Parsing design (AST generation)... Saved backup file: /home/larry/.local/share/OpenSCAD/backups/unsaved- backup-OFwMjSrL.scad Compiling design (CSG Tree generation)... Rendering Polygon Mesh using CGAL... [fast-csg] corefinement mesh union #10: 48274 facets vs. 48274 facets [fast-csg] corefinement mesh union #11: 96548 facets vs. 48274 facets Geometries in cache: 7 Geometry cache size in bytes: 24338664 CGAL Polyhedrons in cache: 0 CGAL cache size in bytes: 0 Total rendering time: 0:00:02.156 Top level object is a 3D object: Facets: 144822 Rendering finished. Caches Flushed Parsing design (AST generation)... Saved backup file: /home/larry/.local/share/OpenSCAD/backups/unsaved- backup-OFwMjSrL.scad Compiling design (CSG Tree generation)... Rendering Polygon Mesh using CGAL... [fast-csg] corefinement mesh difference #2: 476 facets vs. 476 facets [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it seems to have holes. [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it seems to have holes. [fast-csg] corefinement mesh union #12: 1720 facets vs. 476 facets [fast-csg] corefinement mesh union #13: 48274 facets vs. 2196 facets [fast-csg-remesh] Skipping remeshing of patch with 281 faces as it seems to have holes. [fast-csg-remesh] Skipping remeshing of patch with 339 faces as it seems to have holes. [fast-csg-remesh] Remeshed from 67400 to 64816 faces (3 % improvement) [fast-csg] corefinement mesh union #14: 48274 facets vs. 48274 facets [fast-csg] corefinement mesh union #15: 96548 facets vs. 64816 facets [fast-csg-remesh] Skipping remeshing of patch with 603 faces as it seems to have holes. [fast-csg-remesh] Skipping remeshing of patch with 587 faces as it seems to have holes. [fast-csg-remesh] Remeshed from 176804 to 170184 faces (3 % improvement) Geometries in cache: 17 Geometry cache size in bytes: 26321616 CGAL Polyhedrons in cache: 1 CGAL cache size in bytes: 48200 Total rendering time: 0:00:19.915 Top level object is a 3D object: Facets: 170184 Rendering finished. > On Sat, Feb 18, 2023 at 9:20 PM Father Horton > <fatherhorton@gmail.com> wrote: > > I downloaded and ran 2023.2.15, ran each one twice, and flushed the > > cache between runs. > > > > With caps: > > 1 blade: 0:21 both times > > 2 blades: 1:14 both times > > 3 blades: 3:11/3:18 > > > > With caps: > > 1 blade: 1.4 sec/.4 sec > > 2 blades: 1:07 both times > > 3 blades: 3:31/3:35 > > So the dev release does render faster, and I have no idea what > > effect the caps are having. The exponential time increase makes > > sense if OpenSCAD is having to check for collisions of more objects > > or something like that. > > > > _______________________________________________ > > OpenSCAD mailing list > > To unsubscribe send an email to discuss-leave@lists.openscad.org > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org

If there are mesh problems it's in the blades, which have an absurdly high point sampling. I'm not sure what "mesh problems" means or why it would lead to high render times, though. I was going to suggest tweaking the blade design to improve the model, but it doesn't appear necessary. I am guessing that you have lazy unions turned off. Turning lazy unions on seems to help this model enormously---with more advantage than the other speedup stuff. So turn on all the csg things and also turn on lazy unions. On Sun, Feb 19, 2023 at 9:13 AM larry <lar3ry@sasktel.net> wrote: > On Sat, 2023-02-18 at 22:05 -0500, Adrian Mariano wrote: > > I don't know the default configuration: did you turn on all the > > OpenSCAD features that are related to faster rendering? A lot of > > times the new features in the dev versions are off by default. It's > > all the fast-csg things. > > > > For me the 3 blade case rendersin 2022.12.07 in 0.35s. It took 4:14 > > in the stable version of OpenSCAD. > > was that with or without the caps. > I used dev snapshot appimage version 2023.02.15.ai13654 (git 9e8fa3dc5) > > I am wondering if the strange speed results are because of the caps > having mesh problems. > > My results, first run without caps, then with: > > > Caches Flushed > > Parsing design (AST generation)... > Saved backup file: /home/larry/.local/share/OpenSCAD/backups/unsaved- > backup-OFwMjSrL.scad > Compiling design (CSG Tree generation)... > Rendering Polygon Mesh using CGAL... > [fast-csg] corefinement mesh union #10: 48274 facets vs. 48274 facets > [fast-csg] corefinement mesh union #11: 96548 facets vs. 48274 facets > Geometries in cache: 7 > Geometry cache size in bytes: 24338664 > CGAL Polyhedrons in cache: 0 > CGAL cache size in bytes: 0 > Total rendering time: 0:00:02.156 > Top level object is a 3D object: > Facets: 144822 > Rendering finished. > Caches Flushed > > Parsing design (AST generation)... > Saved backup file: /home/larry/.local/share/OpenSCAD/backups/unsaved- > backup-OFwMjSrL.scad > Compiling design (CSG Tree generation)... > Rendering Polygon Mesh using CGAL... > [fast-csg] corefinement mesh difference #2: 476 facets vs. 476 facets > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it > seems to have holes. > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it > seems to have holes. > [fast-csg] corefinement mesh union #12: 1720 facets vs. 476 facets > [fast-csg] corefinement mesh union #13: 48274 facets vs. 2196 facets > [fast-csg-remesh] Skipping remeshing of patch with 281 faces as it > seems to have holes. > [fast-csg-remesh] Skipping remeshing of patch with 339 faces as it > seems to have holes. > [fast-csg-remesh] Remeshed from 67400 to 64816 faces (3 % improvement) > [fast-csg] corefinement mesh union #14: 48274 facets vs. 48274 facets > [fast-csg] corefinement mesh union #15: 96548 facets vs. 64816 facets > [fast-csg-remesh] Skipping remeshing of patch with 603 faces as it > seems to have holes. > [fast-csg-remesh] Skipping remeshing of patch with 587 faces as it > seems to have holes. > [fast-csg-remesh] Remeshed from 176804 to 170184 faces (3 % > improvement) > Geometries in cache: 17 > Geometry cache size in bytes: 26321616 > CGAL Polyhedrons in cache: 1 > CGAL cache size in bytes: 48200 > Total rendering time: 0:00:19.915 > Top level object is a 3D object: > Facets: 170184 > Rendering finished. > > > > On Sat, Feb 18, 2023 at 9:20 PM Father Horton > > <fatherhorton@gmail.com> wrote: > > > I downloaded and ran 2023.2.15, ran each one twice, and flushed the > > > cache between runs. > > > > > > With caps: > > > 1 blade: 0:21 both times > > > 2 blades: 1:14 both times > > > 3 blades: 3:11/3:18 > > > > > > With caps: > > > 1 blade: 1.4 sec/.4 sec > > > 2 blades: 1:07 both times > > > 3 blades: 3:31/3:35 > > > So the dev release does render faster, and I have no idea what > > > effect the caps are having. The exponential time increase makes > > > sense if OpenSCAD is having to check for collisions of more objects > > > or something like that. > > > > > > _______________________________________________ > > > OpenSCAD mailing list > > > To unsubscribe send an email to discuss-leave@lists.openscad.org > > _______________________________________________ > > OpenSCAD mailing list > > To unsubscribe send an email to discuss-leave@lists.openscad.org > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

On Sun, 2023-02-19 at 09:45 -0500, Adrian Mariano wrote: > If there are mesh problems it's in the blades, which have an absurdly > high point sampling.  I'm not sure what "mesh problems" means or why > it would lead to high render times, though.   I was going to suggest > tweaking the blade design to improve the model, but it doesn't appear > necessary.  > > I am guessing that you have lazy unions turned off.  Turning lazy > unions on seems to help this model enormously---with more advantage > than the other speedup stuff.  > > So turn on all the csg things and also turn on lazy unions.  Wow! I already had the fast-csg stuff turned on, but turning on lazy unions REALLY improved things... Caches Flushed Parsing design (AST generation)... Compiling design (CSG Tree generation)... Rendering Polygon Mesh using CGAL... [fast-csg] corefinement mesh difference #1: 476 facets vs. 476 facets [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it seems to have holes. [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it seems to have holes. Geometries in cache: 16 Geometry cache size in bytes: 24561488 CGAL Polyhedrons in cache: 1 CGAL cache size in bytes: 48200 Total rendering time: 0:00:00.204 Top level object is a list of objects: Objects: 5 Rendering finished. > On Sun, Feb 19, 2023 at 9:13 AM larry <lar3ry@sasktel.net> wrote: > > On Sat, 2023-02-18 at 22:05 -0500, Adrian Mariano wrote: > > > I don't know the default configuration:  did you turn on all the > > > OpenSCAD features that are related to faster rendering?  A lot of > > > times the new features in the dev versions are off by default.  > > > It's > > > all the fast-csg things.  > > > > > > For me the 3 blade case rendersin 2022.12.07 in 0.35s.  It took > > > 4:14 > > > in the stable version of OpenSCAD. > > > > was that with or without the caps. > > I used dev snapshot appimage version 2023.02.15.ai13654 (git > > 9e8fa3dc5) > > > > I am wondering if the strange speed results are because of the caps > > having mesh problems. > > > > My results, first run without caps, then with: > > > > > > Caches Flushed > > > > Parsing design (AST generation)... > > Saved backup file: > > /home/larry/.local/share/OpenSCAD/backups/unsaved- > > backup-OFwMjSrL.scad > > Compiling design (CSG Tree generation)... > > Rendering Polygon Mesh using CGAL... > > [fast-csg] corefinement mesh union #10: 48274 facets vs. 48274 > > facets > > [fast-csg] corefinement mesh union #11: 96548 facets vs. 48274 > > facets > > Geometries in cache: 7 > > Geometry cache size in bytes: 24338664 > > CGAL Polyhedrons in cache: 0 > > CGAL cache size in bytes: 0 > > Total rendering time: 0:00:02.156 > > Top level object is a 3D object: > >    Facets:     144822 > > Rendering finished. > > Caches Flushed > > > > Parsing design (AST generation)... > > Saved backup file: > > /home/larry/.local/share/OpenSCAD/backups/unsaved- > > backup-OFwMjSrL.scad > > Compiling design (CSG Tree generation)... > > Rendering Polygon Mesh using CGAL... > > [fast-csg] corefinement mesh difference #2: 476 facets vs. 476 > > facets > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it > > seems to have holes. > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it > > seems to have holes. > > [fast-csg] corefinement mesh union #12: 1720 facets vs. 476 facets > > [fast-csg] corefinement mesh union #13: 48274 facets vs. 2196 > > facets > > [fast-csg-remesh] Skipping remeshing of patch with 281 faces as it > > seems to have holes. > > [fast-csg-remesh] Skipping remeshing of patch with 339 faces as it > > seems to have holes. > > [fast-csg-remesh] Remeshed from 67400 to 64816 faces (3 % > > improvement) > > [fast-csg] corefinement mesh union #14: 48274 facets vs. 48274 > > facets > > [fast-csg] corefinement mesh union #15: 96548 facets vs. 64816 > > facets > > [fast-csg-remesh] Skipping remeshing of patch with 603 faces as it > > seems to have holes. > > [fast-csg-remesh] Skipping remeshing of patch with 587 faces as it > > seems to have holes. > > [fast-csg-remesh] Remeshed from 176804 to 170184 faces (3 % > > improvement) > > Geometries in cache: 17 > > Geometry cache size in bytes: 26321616 > > CGAL Polyhedrons in cache: 1 > > CGAL cache size in bytes: 48200 > > Total rendering time: 0:00:19.915 > > Top level object is a 3D object: > >    Facets:     170184 > > Rendering finished. > > > > > > > On Sat, Feb 18, 2023 at 9:20 PM Father Horton > > > <fatherhorton@gmail.com> wrote: > > > > I downloaded and ran 2023.2.15, ran each one twice, and flushed > > > > the > > > > cache between runs. > > > > > > > > With caps: > > > > 1 blade: 0:21 both times > > > > 2 blades: 1:14 both times > > > > 3 blades: 3:11/3:18 > > > > > > > > With caps: > > > > 1 blade: 1.4 sec/.4 sec > > > > 2 blades: 1:07 both times > > > > 3 blades: 3:31/3:35 > > > > So the dev release does render faster, and I have no idea what > > > > effect the caps are having. The exponential time increase makes > > > > sense if OpenSCAD is having to check for collisions of more > > > > objects > > > > or something like that. > > > > > > > > _______________________________________________ > > > > OpenSCAD mailing list > > > > To unsubscribe send an email to > > > > discuss-leave@lists.openscad.org > > > _______________________________________________ > > > OpenSCAD mailing list > > > To unsubscribe send an email to discuss-leave@lists.openscad.org > > _______________________________________________ > > OpenSCAD mailing list > > To unsubscribe send an email to discuss-leave@lists.openscad.org > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org

Yes but you probably wanted one object, not 5. I.e. you would need to add an explicit union and then it would slow down again. On Sun, 19 Feb 2023 at 16:09, larry <lar3ry@sasktel.net> wrote: > On Sun, 2023-02-19 at 09:45 -0500, Adrian Mariano wrote: > > If there are mesh problems it's in the blades, which have an absurdly > > high point sampling. I'm not sure what "mesh problems" means or why > > it would lead to high render times, though. I was going to suggest > > tweaking the blade design to improve the model, but it doesn't appear > > necessary. > > > > I am guessing that you have lazy unions turned off. Turning lazy > > unions on seems to help this model enormously---with more advantage > > than the other speedup stuff. > > > > So turn on all the csg things and also turn on lazy unions. > > Wow! > I already had the fast-csg stuff turned on, but turning on lazy unions > REALLY improved things... > > > Caches Flushed > > Parsing design (AST generation)... > Compiling design (CSG Tree generation)... > Rendering Polygon Mesh using CGAL... > [fast-csg] corefinement mesh difference #1: 476 facets vs. 476 facets > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it > seems to have holes. > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it > seems to have holes. > Geometries in cache: 16 > Geometry cache size in bytes: 24561488 > CGAL Polyhedrons in cache: 1 > CGAL cache size in bytes: 48200 > Total rendering time: 0:00:00.204 > Top level object is a list of objects: > Objects: 5 > Rendering finished. > > > > On Sun, Feb 19, 2023 at 9:13 AM larry <lar3ry@sasktel.net> wrote: > > > On Sat, 2023-02-18 at 22:05 -0500, Adrian Mariano wrote: > > > > I don't know the default configuration: did you turn on all the > > > > OpenSCAD features that are related to faster rendering? A lot of > > > > times the new features in the dev versions are off by default. > > > > It's > > > > all the fast-csg things. > > > > > > > > For me the 3 blade case rendersin 2022.12.07 in 0.35s. It took > > > > 4:14 > > > > in the stable version of OpenSCAD. > > > > > > was that with or without the caps. > > > I used dev snapshot appimage version 2023.02.15.ai13654 (git > > > 9e8fa3dc5) > > > > > > I am wondering if the strange speed results are because of the caps > > > having mesh problems. > > > > > > My results, first run without caps, then with: > > > > > > > > > Caches Flushed > > > > > > Parsing design (AST generation)... > > > Saved backup file: > > > /home/larry/.local/share/OpenSCAD/backups/unsaved- > > > backup-OFwMjSrL.scad > > > Compiling design (CSG Tree generation)... > > > Rendering Polygon Mesh using CGAL... > > > [fast-csg] corefinement mesh union #10: 48274 facets vs. 48274 > > > facets > > > [fast-csg] corefinement mesh union #11: 96548 facets vs. 48274 > > > facets > > > Geometries in cache: 7 > > > Geometry cache size in bytes: 24338664 > > > CGAL Polyhedrons in cache: 0 > > > CGAL cache size in bytes: 0 > > > Total rendering time: 0:00:02.156 > > > Top level object is a 3D object: > > > Facets: 144822 > > > Rendering finished. > > > Caches Flushed > > > > > > Parsing design (AST generation)... > > > Saved backup file: > > > /home/larry/.local/share/OpenSCAD/backups/unsaved- > > > backup-OFwMjSrL.scad > > > Compiling design (CSG Tree generation)... > > > Rendering Polygon Mesh using CGAL... > > > [fast-csg] corefinement mesh difference #2: 476 facets vs. 476 > > > facets > > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it > > > seems to have holes. > > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it > > > seems to have holes. > > > [fast-csg] corefinement mesh union #12: 1720 facets vs. 476 facets > > > [fast-csg] corefinement mesh union #13: 48274 facets vs. 2196 > > > facets > > > [fast-csg-remesh] Skipping remeshing of patch with 281 faces as it > > > seems to have holes. > > > [fast-csg-remesh] Skipping remeshing of patch with 339 faces as it > > > seems to have holes. > > > [fast-csg-remesh] Remeshed from 67400 to 64816 faces (3 % > > > improvement) > > > [fast-csg] corefinement mesh union #14: 48274 facets vs. 48274 > > > facets > > > [fast-csg] corefinement mesh union #15: 96548 facets vs. 64816 > > > facets > > > [fast-csg-remesh] Skipping remeshing of patch with 603 faces as it > > > seems to have holes. > > > [fast-csg-remesh] Skipping remeshing of patch with 587 faces as it > > > seems to have holes. > > > [fast-csg-remesh] Remeshed from 176804 to 170184 faces (3 % > > > improvement) > > > Geometries in cache: 17 > > > Geometry cache size in bytes: 26321616 > > > CGAL Polyhedrons in cache: 1 > > > CGAL cache size in bytes: 48200 > > > Total rendering time: 0:00:19.915 > > > Top level object is a 3D object: > > > Facets: 170184 > > > Rendering finished. > > > > > > > > > > On Sat, Feb 18, 2023 at 9:20 PM Father Horton > > > > <fatherhorton@gmail.com> wrote: > > > > > I downloaded and ran 2023.2.15, ran each one twice, and flushed > > > > > the > > > > > cache between runs. > > > > > > > > > > With caps: > > > > > 1 blade: 0:21 both times > > > > > 2 blades: 1:14 both times > > > > > 3 blades: 3:11/3:18 > > > > > > > > > > With caps: > > > > > 1 blade: 1.4 sec/.4 sec > > > > > 2 blades: 1:07 both times > > > > > 3 blades: 3:31/3:35 > > > > > So the dev release does render faster, and I have no idea what > > > > > effect the caps are having. The exponential time increase makes > > > > > sense if OpenSCAD is having to check for collisions of more > > > > > objects > > > > > or something like that. > > > > > > > > > > _______________________________________________ > > > > > OpenSCAD mailing list > > > > > To unsubscribe send an email to > > > > > discuss-leave@lists.openscad.org > > > > _______________________________________________ > > > > OpenSCAD mailing list > > > > To unsubscribe send an email to discuss-leave@lists.openscad.org > > > _______________________________________________ > > > OpenSCAD mailing list > > > To unsubscribe send an email to discuss-leave@lists.openscad.org > > _______________________________________________ > > OpenSCAD mailing list > > To unsubscribe send an email to discuss-leave@lists.openscad.org > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

Hmmm. That does increase render to 22s for me. What happens if you export to STL a model with several objects? Is it a problem? I just tried doing it and loaded the file into my slicer and the slicer produced a reasonable looking result. Nothing seems to have gone wrong. On Sun, Feb 19, 2023 at 11:12 AM nop head <nop.head@gmail.com> wrote: > Yes but you probably wanted one object, not 5. I.e. you would need to add > an explicit union and then it would slow down again. > > On Sun, 19 Feb 2023 at 16:09, larry <lar3ry@sasktel.net> wrote: > >> On Sun, 2023-02-19 at 09:45 -0500, Adrian Mariano wrote: >> > If there are mesh problems it's in the blades, which have an absurdly >> > high point sampling. I'm not sure what "mesh problems" means or why >> > it would lead to high render times, though. I was going to suggest >> > tweaking the blade design to improve the model, but it doesn't appear >> > necessary. >> > >> > I am guessing that you have lazy unions turned off. Turning lazy >> > unions on seems to help this model enormously---with more advantage >> > than the other speedup stuff. >> > >> > So turn on all the csg things and also turn on lazy unions. >> >> Wow! >> I already had the fast-csg stuff turned on, but turning on lazy unions >> REALLY improved things... >> >> >> Caches Flushed >> >> Parsing design (AST generation)... >> Compiling design (CSG Tree generation)... >> Rendering Polygon Mesh using CGAL... >> [fast-csg] corefinement mesh difference #1: 476 facets vs. 476 facets >> [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >> seems to have holes. >> [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >> seems to have holes. >> Geometries in cache: 16 >> Geometry cache size in bytes: 24561488 >> CGAL Polyhedrons in cache: 1 >> CGAL cache size in bytes: 48200 >> Total rendering time: 0:00:00.204 >> Top level object is a list of objects: >> Objects: 5 >> Rendering finished. >> >> >> > On Sun, Feb 19, 2023 at 9:13 AM larry <lar3ry@sasktel.net> wrote: >> > > On Sat, 2023-02-18 at 22:05 -0500, Adrian Mariano wrote: >> > > > I don't know the default configuration: did you turn on all the >> > > > OpenSCAD features that are related to faster rendering? A lot of >> > > > times the new features in the dev versions are off by default. >> > > > It's >> > > > all the fast-csg things. >> > > > >> > > > For me the 3 blade case rendersin 2022.12.07 in 0.35s. It took >> > > > 4:14 >> > > > in the stable version of OpenSCAD. >> > > >> > > was that with or without the caps. >> > > I used dev snapshot appimage version 2023.02.15.ai13654 (git >> > > 9e8fa3dc5) >> > > >> > > I am wondering if the strange speed results are because of the caps >> > > having mesh problems. >> > > >> > > My results, first run without caps, then with: >> > > >> > > >> > > Caches Flushed >> > > >> > > Parsing design (AST generation)... >> > > Saved backup file: >> > > /home/larry/.local/share/OpenSCAD/backups/unsaved- >> > > backup-OFwMjSrL.scad >> > > Compiling design (CSG Tree generation)... >> > > Rendering Polygon Mesh using CGAL... >> > > [fast-csg] corefinement mesh union #10: 48274 facets vs. 48274 >> > > facets >> > > [fast-csg] corefinement mesh union #11: 96548 facets vs. 48274 >> > > facets >> > > Geometries in cache: 7 >> > > Geometry cache size in bytes: 24338664 >> > > CGAL Polyhedrons in cache: 0 >> > > CGAL cache size in bytes: 0 >> > > Total rendering time: 0:00:02.156 >> > > Top level object is a 3D object: >> > > Facets: 144822 >> > > Rendering finished. >> > > Caches Flushed >> > > >> > > Parsing design (AST generation)... >> > > Saved backup file: >> > > /home/larry/.local/share/OpenSCAD/backups/unsaved- >> > > backup-OFwMjSrL.scad >> > > Compiling design (CSG Tree generation)... >> > > Rendering Polygon Mesh using CGAL... >> > > [fast-csg] corefinement mesh difference #2: 476 facets vs. 476 >> > > facets >> > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >> > > seems to have holes. >> > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >> > > seems to have holes. >> > > [fast-csg] corefinement mesh union #12: 1720 facets vs. 476 facets >> > > [fast-csg] corefinement mesh union #13: 48274 facets vs. 2196 >> > > facets >> > > [fast-csg-remesh] Skipping remeshing of patch with 281 faces as it >> > > seems to have holes. >> > > [fast-csg-remesh] Skipping remeshing of patch with 339 faces as it >> > > seems to have holes. >> > > [fast-csg-remesh] Remeshed from 67400 to 64816 faces (3 % >> > > improvement) >> > > [fast-csg] corefinement mesh union #14: 48274 facets vs. 48274 >> > > facets >> > > [fast-csg] corefinement mesh union #15: 96548 facets vs. 64816 >> > > facets >> > > [fast-csg-remesh] Skipping remeshing of patch with 603 faces as it >> > > seems to have holes. >> > > [fast-csg-remesh] Skipping remeshing of patch with 587 faces as it >> > > seems to have holes. >> > > [fast-csg-remesh] Remeshed from 176804 to 170184 faces (3 % >> > > improvement) >> > > Geometries in cache: 17 >> > > Geometry cache size in bytes: 26321616 >> > > CGAL Polyhedrons in cache: 1 >> > > CGAL cache size in bytes: 48200 >> > > Total rendering time: 0:00:19.915 >> > > Top level object is a 3D object: >> > > Facets: 170184 >> > > Rendering finished. >> > > >> > > >> > > > On Sat, Feb 18, 2023 at 9:20 PM Father Horton >> > > > <fatherhorton@gmail.com> wrote: >> > > > > I downloaded and ran 2023.2.15, ran each one twice, and flushed >> > > > > the >> > > > > cache between runs. >> > > > > >> > > > > With caps: >> > > > > 1 blade: 0:21 both times >> > > > > 2 blades: 1:14 both times >> > > > > 3 blades: 3:11/3:18 >> > > > > >> > > > > With caps: >> > > > > 1 blade: 1.4 sec/.4 sec >> > > > > 2 blades: 1:07 both times >> > > > > 3 blades: 3:31/3:35 >> > > > > So the dev release does render faster, and I have no idea what >> > > > > effect the caps are having. The exponential time increase makes >> > > > > sense if OpenSCAD is having to check for collisions of more >> > > > > objects >> > > > > or something like that. >> > > > > >> > > > > _______________________________________________ >> > > > > OpenSCAD mailing list >> > > > > To unsubscribe send an email to >> > > > > discuss-leave@lists.openscad.org >> > > > _______________________________________________ >> > > > OpenSCAD mailing list >> > > > To unsubscribe send an email to discuss-leave@lists.openscad.org >> > > _______________________________________________ >> > > OpenSCAD mailing list >> > > To unsubscribe send an email to discuss-leave@lists.openscad.org >> > _______________________________________________ >> > OpenSCAD mailing list >> > To unsubscribe send an email to discuss-leave@lists.openscad.org >> _______________________________________________ >> OpenSCAD mailing list >> To unsubscribe send an email to discuss-leave@lists.openscad.org >> > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

The slicer I use will exor overlapping objects, rather than unioning them. Also you won't be able to import the STL back into OpenSCAD if it has overlapping objects. On Sun, 19 Feb 2023 at 17:49, Adrian Mariano <avm4@cornell.edu> wrote: > Hmmm. That does increase render to 22s for me. What happens if you > export to STL a model with several objects? Is it a problem? I just tried > doing it and loaded the file into my slicer and the slicer produced a > reasonable looking result. Nothing seems to have gone wrong. > > On Sun, Feb 19, 2023 at 11:12 AM nop head <nop.head@gmail.com> wrote: > >> Yes but you probably wanted one object, not 5. I.e. you would need to add >> an explicit union and then it would slow down again. >> >> On Sun, 19 Feb 2023 at 16:09, larry <lar3ry@sasktel.net> wrote: >> >>> On Sun, 2023-02-19 at 09:45 -0500, Adrian Mariano wrote: >>> > If there are mesh problems it's in the blades, which have an absurdly >>> > high point sampling. I'm not sure what "mesh problems" means or why >>> > it would lead to high render times, though. I was going to suggest >>> > tweaking the blade design to improve the model, but it doesn't appear >>> > necessary. >>> > >>> > I am guessing that you have lazy unions turned off. Turning lazy >>> > unions on seems to help this model enormously---with more advantage >>> > than the other speedup stuff. >>> > >>> > So turn on all the csg things and also turn on lazy unions. >>> >>> Wow! >>> I already had the fast-csg stuff turned on, but turning on lazy unions >>> REALLY improved things... >>> >>> >>> Caches Flushed >>> >>> Parsing design (AST generation)... >>> Compiling design (CSG Tree generation)... >>> Rendering Polygon Mesh using CGAL... >>> [fast-csg] corefinement mesh difference #1: 476 facets vs. 476 facets >>> [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>> seems to have holes. >>> [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>> seems to have holes. >>> Geometries in cache: 16 >>> Geometry cache size in bytes: 24561488 >>> CGAL Polyhedrons in cache: 1 >>> CGAL cache size in bytes: 48200 >>> Total rendering time: 0:00:00.204 >>> Top level object is a list of objects: >>> Objects: 5 >>> Rendering finished. >>> >>> >>> > On Sun, Feb 19, 2023 at 9:13 AM larry <lar3ry@sasktel.net> wrote: >>> > > On Sat, 2023-02-18 at 22:05 -0500, Adrian Mariano wrote: >>> > > > I don't know the default configuration: did you turn on all the >>> > > > OpenSCAD features that are related to faster rendering? A lot of >>> > > > times the new features in the dev versions are off by default. >>> > > > It's >>> > > > all the fast-csg things. >>> > > > >>> > > > For me the 3 blade case rendersin 2022.12.07 in 0.35s. It took >>> > > > 4:14 >>> > > > in the stable version of OpenSCAD. >>> > > >>> > > was that with or without the caps. >>> > > I used dev snapshot appimage version 2023.02.15.ai13654 (git >>> > > 9e8fa3dc5) >>> > > >>> > > I am wondering if the strange speed results are because of the caps >>> > > having mesh problems. >>> > > >>> > > My results, first run without caps, then with: >>> > > >>> > > >>> > > Caches Flushed >>> > > >>> > > Parsing design (AST generation)... >>> > > Saved backup file: >>> > > /home/larry/.local/share/OpenSCAD/backups/unsaved- >>> > > backup-OFwMjSrL.scad >>> > > Compiling design (CSG Tree generation)... >>> > > Rendering Polygon Mesh using CGAL... >>> > > [fast-csg] corefinement mesh union #10: 48274 facets vs. 48274 >>> > > facets >>> > > [fast-csg] corefinement mesh union #11: 96548 facets vs. 48274 >>> > > facets >>> > > Geometries in cache: 7 >>> > > Geometry cache size in bytes: 24338664 >>> > > CGAL Polyhedrons in cache: 0 >>> > > CGAL cache size in bytes: 0 >>> > > Total rendering time: 0:00:02.156 >>> > > Top level object is a 3D object: >>> > > Facets: 144822 >>> > > Rendering finished. >>> > > Caches Flushed >>> > > >>> > > Parsing design (AST generation)... >>> > > Saved backup file: >>> > > /home/larry/.local/share/OpenSCAD/backups/unsaved- >>> > > backup-OFwMjSrL.scad >>> > > Compiling design (CSG Tree generation)... >>> > > Rendering Polygon Mesh using CGAL... >>> > > [fast-csg] corefinement mesh difference #2: 476 facets vs. 476 >>> > > facets >>> > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>> > > seems to have holes. >>> > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>> > > seems to have holes. >>> > > [fast-csg] corefinement mesh union #12: 1720 facets vs. 476 facets >>> > > [fast-csg] corefinement mesh union #13: 48274 facets vs. 2196 >>> > > facets >>> > > [fast-csg-remesh] Skipping remeshing of patch with 281 faces as it >>> > > seems to have holes. >>> > > [fast-csg-remesh] Skipping remeshing of patch with 339 faces as it >>> > > seems to have holes. >>> > > [fast-csg-remesh] Remeshed from 67400 to 64816 faces (3 % >>> > > improvement) >>> > > [fast-csg] corefinement mesh union #14: 48274 facets vs. 48274 >>> > > facets >>> > > [fast-csg] corefinement mesh union #15: 96548 facets vs. 64816 >>> > > facets >>> > > [fast-csg-remesh] Skipping remeshing of patch with 603 faces as it >>> > > seems to have holes. >>> > > [fast-csg-remesh] Skipping remeshing of patch with 587 faces as it >>> > > seems to have holes. >>> > > [fast-csg-remesh] Remeshed from 176804 to 170184 faces (3 % >>> > > improvement) >>> > > Geometries in cache: 17 >>> > > Geometry cache size in bytes: 26321616 >>> > > CGAL Polyhedrons in cache: 1 >>> > > CGAL cache size in bytes: 48200 >>> > > Total rendering time: 0:00:19.915 >>> > > Top level object is a 3D object: >>> > > Facets: 170184 >>> > > Rendering finished. >>> > > >>> > > >>> > > > On Sat, Feb 18, 2023 at 9:20 PM Father Horton >>> > > > <fatherhorton@gmail.com> wrote: >>> > > > > I downloaded and ran 2023.2.15, ran each one twice, and flushed >>> > > > > the >>> > > > > cache between runs. >>> > > > > >>> > > > > With caps: >>> > > > > 1 blade: 0:21 both times >>> > > > > 2 blades: 1:14 both times >>> > > > > 3 blades: 3:11/3:18 >>> > > > > >>> > > > > With caps: >>> > > > > 1 blade: 1.4 sec/.4 sec >>> > > > > 2 blades: 1:07 both times >>> > > > > 3 blades: 3:31/3:35 >>> > > > > So the dev release does render faster, and I have no idea what >>> > > > > effect the caps are having. The exponential time increase makes >>> > > > > sense if OpenSCAD is having to check for collisions of more >>> > > > > objects >>> > > > > or something like that. >>> > > > > >>> > > > > _______________________________________________ >>> > > > > OpenSCAD mailing list >>> > > > > To unsubscribe send an email to >>> > > > > discuss-leave@lists.openscad.org >>> > > > _______________________________________________ >>> > > > OpenSCAD mailing list >>> > > > To unsubscribe send an email to discuss-leave@lists.openscad.org >>> > > _______________________________________________ >>> > > OpenSCAD mailing list >>> > > To unsubscribe send an email to discuss-leave@lists.openscad.org >>> > _______________________________________________ >>> > OpenSCAD mailing list >>> > To unsubscribe send an email to discuss-leave@lists.openscad.org >>> _______________________________________________ >>> OpenSCAD mailing list >>> To unsubscribe send an email to discuss-leave@lists.openscad.org >>> >> _______________________________________________ >> OpenSCAD mailing list >> To unsubscribe send an email to discuss-leave@lists.openscad.org >> > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

I'm using PrusaSlicer, which seems to union the objects in the STL, based on a simple test case that contained a cube and a cylinder. On Sun, Feb 19, 2023 at 12:53 PM nop head <nop.head@gmail.com> wrote: > The slicer I use will exor overlapping objects, rather than unioning them. > Also you won't be able to import the STL back into OpenSCAD if it has > overlapping objects. > > On Sun, 19 Feb 2023 at 17:49, Adrian Mariano <avm4@cornell.edu> wrote: > >> Hmmm. That does increase render to 22s for me. What happens if you >> export to STL a model with several objects? Is it a problem? I just tried >> doing it and loaded the file into my slicer and the slicer produced a >> reasonable looking result. Nothing seems to have gone wrong. >> >> On Sun, Feb 19, 2023 at 11:12 AM nop head <nop.head@gmail.com> wrote: >> >>> Yes but you probably wanted one object, not 5. I.e. you would need to >>> add an explicit union and then it would slow down again. >>> >>> On Sun, 19 Feb 2023 at 16:09, larry <lar3ry@sasktel.net> wrote: >>> >>>> On Sun, 2023-02-19 at 09:45 -0500, Adrian Mariano wrote: >>>> > If there are mesh problems it's in the blades, which have an absurdly >>>> > high point sampling. I'm not sure what "mesh problems" means or why >>>> > it would lead to high render times, though. I was going to suggest >>>> > tweaking the blade design to improve the model, but it doesn't appear >>>> > necessary. >>>> > >>>> > I am guessing that you have lazy unions turned off. Turning lazy >>>> > unions on seems to help this model enormously---with more advantage >>>> > than the other speedup stuff. >>>> > >>>> > So turn on all the csg things and also turn on lazy unions. >>>> >>>> Wow! >>>> I already had the fast-csg stuff turned on, but turning on lazy unions >>>> REALLY improved things... >>>> >>>> >>>> Caches Flushed >>>> >>>> Parsing design (AST generation)... >>>> Compiling design (CSG Tree generation)... >>>> Rendering Polygon Mesh using CGAL... >>>> [fast-csg] corefinement mesh difference #1: 476 facets vs. 476 facets >>>> [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>>> seems to have holes. >>>> [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>>> seems to have holes. >>>> Geometries in cache: 16 >>>> Geometry cache size in bytes: 24561488 >>>> CGAL Polyhedrons in cache: 1 >>>> CGAL cache size in bytes: 48200 >>>> Total rendering time: 0:00:00.204 >>>> Top level object is a list of objects: >>>> Objects: 5 >>>> Rendering finished. >>>> >>>> >>>> > On Sun, Feb 19, 2023 at 9:13 AM larry <lar3ry@sasktel.net> wrote: >>>> > > On Sat, 2023-02-18 at 22:05 -0500, Adrian Mariano wrote: >>>> > > > I don't know the default configuration: did you turn on all the >>>> > > > OpenSCAD features that are related to faster rendering? A lot of >>>> > > > times the new features in the dev versions are off by default. >>>> > > > It's >>>> > > > all the fast-csg things. >>>> > > > >>>> > > > For me the 3 blade case rendersin 2022.12.07 in 0.35s. It took >>>> > > > 4:14 >>>> > > > in the stable version of OpenSCAD. >>>> > > >>>> > > was that with or without the caps. >>>> > > I used dev snapshot appimage version 2023.02.15.ai13654 (git >>>> > > 9e8fa3dc5) >>>> > > >>>> > > I am wondering if the strange speed results are because of the caps >>>> > > having mesh problems. >>>> > > >>>> > > My results, first run without caps, then with: >>>> > > >>>> > > >>>> > > Caches Flushed >>>> > > >>>> > > Parsing design (AST generation)... >>>> > > Saved backup file: >>>> > > /home/larry/.local/share/OpenSCAD/backups/unsaved- >>>> > > backup-OFwMjSrL.scad >>>> > > Compiling design (CSG Tree generation)... >>>> > > Rendering Polygon Mesh using CGAL... >>>> > > [fast-csg] corefinement mesh union #10: 48274 facets vs. 48274 >>>> > > facets >>>> > > [fast-csg] corefinement mesh union #11: 96548 facets vs. 48274 >>>> > > facets >>>> > > Geometries in cache: 7 >>>> > > Geometry cache size in bytes: 24338664 >>>> > > CGAL Polyhedrons in cache: 0 >>>> > > CGAL cache size in bytes: 0 >>>> > > Total rendering time: 0:00:02.156 >>>> > > Top level object is a 3D object: >>>> > > Facets: 144822 >>>> > > Rendering finished. >>>> > > Caches Flushed >>>> > > >>>> > > Parsing design (AST generation)... >>>> > > Saved backup file: >>>> > > /home/larry/.local/share/OpenSCAD/backups/unsaved- >>>> > > backup-OFwMjSrL.scad >>>> > > Compiling design (CSG Tree generation)... >>>> > > Rendering Polygon Mesh using CGAL... >>>> > > [fast-csg] corefinement mesh difference #2: 476 facets vs. 476 >>>> > > facets >>>> > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>>> > > seems to have holes. >>>> > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>>> > > seems to have holes. >>>> > > [fast-csg] corefinement mesh union #12: 1720 facets vs. 476 facets >>>> > > [fast-csg] corefinement mesh union #13: 48274 facets vs. 2196 >>>> > > facets >>>> > > [fast-csg-remesh] Skipping remeshing of patch with 281 faces as it >>>> > > seems to have holes. >>>> > > [fast-csg-remesh] Skipping remeshing of patch with 339 faces as it >>>> > > seems to have holes. >>>> > > [fast-csg-remesh] Remeshed from 67400 to 64816 faces (3 % >>>> > > improvement) >>>> > > [fast-csg] corefinement mesh union #14: 48274 facets vs. 48274 >>>> > > facets >>>> > > [fast-csg] corefinement mesh union #15: 96548 facets vs. 64816 >>>> > > facets >>>> > > [fast-csg-remesh] Skipping remeshing of patch with 603 faces as it >>>> > > seems to have holes. >>>> > > [fast-csg-remesh] Skipping remeshing of patch with 587 faces as it >>>> > > seems to have holes. >>>> > > [fast-csg-remesh] Remeshed from 176804 to 170184 faces (3 % >>>> > > improvement) >>>> > > Geometries in cache: 17 >>>> > > Geometry cache size in bytes: 26321616 >>>> > > CGAL Polyhedrons in cache: 1 >>>> > > CGAL cache size in bytes: 48200 >>>> > > Total rendering time: 0:00:19.915 >>>> > > Top level object is a 3D object: >>>> > > Facets: 170184 >>>> > > Rendering finished. >>>> > > >>>> > > >>>> > > > On Sat, Feb 18, 2023 at 9:20 PM Father Horton >>>> > > > <fatherhorton@gmail.com> wrote: >>>> > > > > I downloaded and ran 2023.2.15, ran each one twice, and flushed >>>> > > > > the >>>> > > > > cache between runs. >>>> > > > > >>>> > > > > With caps: >>>> > > > > 1 blade: 0:21 both times >>>> > > > > 2 blades: 1:14 both times >>>> > > > > 3 blades: 3:11/3:18 >>>> > > > > >>>> > > > > With caps: >>>> > > > > 1 blade: 1.4 sec/.4 sec >>>> > > > > 2 blades: 1:07 both times >>>> > > > > 3 blades: 3:31/3:35 >>>> > > > > So the dev release does render faster, and I have no idea what >>>> > > > > effect the caps are having. The exponential time increase makes >>>> > > > > sense if OpenSCAD is having to check for collisions of more >>>> > > > > objects >>>> > > > > or something like that. >>>> > > > > >>>> > > > > _______________________________________________ >>>> > > > > OpenSCAD mailing list >>>> > > > > To unsubscribe send an email to >>>> > > > > discuss-leave@lists.openscad.org >>>> > > > _______________________________________________ >>>> > > > OpenSCAD mailing list >>>> > > > To unsubscribe send an email to discuss-leave@lists.openscad.org >>>> > > _______________________________________________ >>>> > > OpenSCAD mailing list >>>> > > To unsubscribe send an email to discuss-leave@lists.openscad.org >>>> > _______________________________________________ >>>> > OpenSCAD mailing list >>>> > To unsubscribe send an email to discuss-leave@lists.openscad.org >>>> _______________________________________________ >>>> OpenSCAD mailing list >>>> To unsubscribe send an email to discuss-leave@lists.openscad.org >>>> >>> _______________________________________________ >>> OpenSCAD mailing list >>> To unsubscribe send an email to discuss-leave@lists.openscad.org >>> >> _______________________________________________ >> OpenSCAD mailing list >> To unsubscribe send an email to discuss-leave@lists.openscad.org >> > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

I use Skienforge which hasn't been updated for about a decade now and I still use Win7 so stuck on an old version of OpenSCAD as well. On Sun, 19 Feb 2023 at 19:30, Adrian Mariano <avm4@cornell.edu> wrote: > I'm using PrusaSlicer, which seems to union the objects in the STL, based > on a simple test case that contained a cube and a cylinder. > > On Sun, Feb 19, 2023 at 12:53 PM nop head <nop.head@gmail.com> wrote: > >> The slicer I use will exor overlapping objects, rather than unioning >> them. Also you won't be able to import the STL back into OpenSCAD if it has >> overlapping objects. >> >> On Sun, 19 Feb 2023 at 17:49, Adrian Mariano <avm4@cornell.edu> wrote: >> >>> Hmmm. That does increase render to 22s for me. What happens if you >>> export to STL a model with several objects? Is it a problem? I just tried >>> doing it and loaded the file into my slicer and the slicer produced a >>> reasonable looking result. Nothing seems to have gone wrong. >>> >>> On Sun, Feb 19, 2023 at 11:12 AM nop head <nop.head@gmail.com> wrote: >>> >>>> Yes but you probably wanted one object, not 5. I.e. you would need to >>>> add an explicit union and then it would slow down again. >>>> >>>> On Sun, 19 Feb 2023 at 16:09, larry <lar3ry@sasktel.net> wrote: >>>> >>>>> On Sun, 2023-02-19 at 09:45 -0500, Adrian Mariano wrote: >>>>> > If there are mesh problems it's in the blades, which have an absurdly >>>>> > high point sampling. I'm not sure what "mesh problems" means or why >>>>> > it would lead to high render times, though. I was going to suggest >>>>> > tweaking the blade design to improve the model, but it doesn't appear >>>>> > necessary. >>>>> > >>>>> > I am guessing that you have lazy unions turned off. Turning lazy >>>>> > unions on seems to help this model enormously---with more advantage >>>>> > than the other speedup stuff. >>>>> > >>>>> > So turn on all the csg things and also turn on lazy unions. >>>>> >>>>> Wow! >>>>> I already had the fast-csg stuff turned on, but turning on lazy unions >>>>> REALLY improved things... >>>>> >>>>> >>>>> Caches Flushed >>>>> >>>>> Parsing design (AST generation)... >>>>> Compiling design (CSG Tree generation)... >>>>> Rendering Polygon Mesh using CGAL... >>>>> [fast-csg] corefinement mesh difference #1: 476 facets vs. 476 facets >>>>> [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>>>> seems to have holes. >>>>> [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>>>> seems to have holes. >>>>> Geometries in cache: 16 >>>>> Geometry cache size in bytes: 24561488 >>>>> CGAL Polyhedrons in cache: 1 >>>>> CGAL cache size in bytes: 48200 >>>>> Total rendering time: 0:00:00.204 >>>>> Top level object is a list of objects: >>>>> Objects: 5 >>>>> Rendering finished. >>>>> >>>>> >>>>> > On Sun, Feb 19, 2023 at 9:13 AM larry <lar3ry@sasktel.net> wrote: >>>>> > > On Sat, 2023-02-18 at 22:05 -0500, Adrian Mariano wrote: >>>>> > > > I don't know the default configuration: did you turn on all the >>>>> > > > OpenSCAD features that are related to faster rendering? A lot of >>>>> > > > times the new features in the dev versions are off by default. >>>>> > > > It's >>>>> > > > all the fast-csg things. >>>>> > > > >>>>> > > > For me the 3 blade case rendersin 2022.12.07 in 0.35s. It took >>>>> > > > 4:14 >>>>> > > > in the stable version of OpenSCAD. >>>>> > > >>>>> > > was that with or without the caps. >>>>> > > I used dev snapshot appimage version 2023.02.15.ai13654 (git >>>>> > > 9e8fa3dc5) >>>>> > > >>>>> > > I am wondering if the strange speed results are because of the caps >>>>> > > having mesh problems. >>>>> > > >>>>> > > My results, first run without caps, then with: >>>>> > > >>>>> > > >>>>> > > Caches Flushed >>>>> > > >>>>> > > Parsing design (AST generation)... >>>>> > > Saved backup file: >>>>> > > /home/larry/.local/share/OpenSCAD/backups/unsaved- >>>>> > > backup-OFwMjSrL.scad >>>>> > > Compiling design (CSG Tree generation)... >>>>> > > Rendering Polygon Mesh using CGAL... >>>>> > > [fast-csg] corefinement mesh union #10: 48274 facets vs. 48274 >>>>> > > facets >>>>> > > [fast-csg] corefinement mesh union #11: 96548 facets vs. 48274 >>>>> > > facets >>>>> > > Geometries in cache: 7 >>>>> > > Geometry cache size in bytes: 24338664 >>>>> > > CGAL Polyhedrons in cache: 0 >>>>> > > CGAL cache size in bytes: 0 >>>>> > > Total rendering time: 0:00:02.156 >>>>> > > Top level object is a 3D object: >>>>> > > Facets: 144822 >>>>> > > Rendering finished. >>>>> > > Caches Flushed >>>>> > > >>>>> > > Parsing design (AST generation)... >>>>> > > Saved backup file: >>>>> > > /home/larry/.local/share/OpenSCAD/backups/unsaved- >>>>> > > backup-OFwMjSrL.scad >>>>> > > Compiling design (CSG Tree generation)... >>>>> > > Rendering Polygon Mesh using CGAL... >>>>> > > [fast-csg] corefinement mesh difference #2: 476 facets vs. 476 >>>>> > > facets >>>>> > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>>>> > > seems to have holes. >>>>> > > [fast-csg-remesh] Skipping remeshing of patch with 120 faces as it >>>>> > > seems to have holes. >>>>> > > [fast-csg] corefinement mesh union #12: 1720 facets vs. 476 facets >>>>> > > [fast-csg] corefinement mesh union #13: 48274 facets vs. 2196 >>>>> > > facets >>>>> > > [fast-csg-remesh] Skipping remeshing of patch with 281 faces as it >>>>> > > seems to have holes. >>>>> > > [fast-csg-remesh] Skipping remeshing of patch with 339 faces as it >>>>> > > seems to have holes. >>>>> > > [fast-csg-remesh] Remeshed from 67400 to 64816 faces (3 % >>>>> > > improvement) >>>>> > > [fast-csg] corefinement mesh union #14: 48274 facets vs. 48274 >>>>> > > facets >>>>> > > [fast-csg] corefinement mesh union #15: 96548 facets vs. 64816 >>>>> > > facets >>>>> > > [fast-csg-remesh] Skipping remeshing of patch with 603 faces as it >>>>> > > seems to have holes. >>>>> > > [fast-csg-remesh] Skipping remeshing of patch with 587 faces as it >>>>> > > seems to have holes. >>>>> > > [fast-csg-remesh] Remeshed from 176804 to 170184 faces (3 % >>>>> > > improvement) >>>>> > > Geometries in cache: 17 >>>>> > > Geometry cache size in bytes: 26321616 >>>>> > > CGAL Polyhedrons in cache: 1 >>>>> > > CGAL cache size in bytes: 48200 >>>>> > > Total rendering time: 0:00:19.915 >>>>> > > Top level object is a 3D object: >>>>> > > Facets: 170184 >>>>> > > Rendering finished. >>>>> > > >>>>> > > >>>>> > > > On Sat, Feb 18, 2023 at 9:20 PM Father Horton >>>>> > > > <fatherhorton@gmail.com> wrote: >>>>> > > > > I downloaded and ran 2023.2.15, ran each one twice, and flushed >>>>> > > > > the >>>>> > > > > cache between runs. >>>>> > > > > >>>>> > > > > With caps: >>>>> > > > > 1 blade: 0:21 both times >>>>> > > > > 2 blades: 1:14 both times >>>>> > > > > 3 blades: 3:11/3:18 >>>>> > > > > >>>>> > > > > With caps: >>>>> > > > > 1 blade: 1.4 sec/.4 sec >>>>> > > > > 2 blades: 1:07 both times >>>>> > > > > 3 blades: 3:31/3:35 >>>>> > > > > So the dev release does render faster, and I have no idea what >>>>> > > > > effect the caps are having. The exponential time increase makes >>>>> > > > > sense if OpenSCAD is having to check for collisions of more >>>>> > > > > objects >>>>> > > > > or something like that. >>>>> > > > > >>>>> > > > > _______________________________________________ >>>>> > > > > OpenSCAD mailing list >>>>> > > > > To unsubscribe send an email to >>>>> > > > > discuss-leave@lists.openscad.org >>>>> > > > _______________________________________________ >>>>> > > > OpenSCAD mailing list >>>>> > > > To unsubscribe send an email to discuss-leave@lists.openscad.org >>>>> > > _______________________________________________ >>>>> > > OpenSCAD mailing list >>>>> > > To unsubscribe send an email to discuss-leave@lists.openscad.org >>>>> > _______________________________________________ >>>>> > OpenSCAD mailing list >>>>> > To unsubscribe send an email to discuss-leave@lists.openscad.org >>>>> _______________________________________________ >>>>> OpenSCAD mailing list >>>>> To unsubscribe send an email to discuss-leave@lists.openscad.org >>>>> >>>> _______________________________________________ >>>> OpenSCAD mailing list >>>> To unsubscribe send an email to discuss-leave@lists.openscad.org >>>> >>> _______________________________________________ >>> OpenSCAD mailing list >>> To unsubscribe send an email to discuss-leave@lists.openscad.org >>> >> _______________________________________________ >> OpenSCAD mailing list >> To unsubscribe send an email to discuss-leave@lists.openscad.org >> > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >