02:50 <gbruno> [github] kintel pushed 2 modifications (Make createSurfaceMeshFromPolySet() return shared_ptr to align with similar functions) https://github.com/openscad/openscad/commit/279660eb2a0ca24786ed17cc245af9735f6920f4

02:50 <gbruno> [github] kintel synchronize pull request #5872 (Refactoring to clean up and align CGAL geometry conversions a bit.) https://github.com/openscad/openscad/pull/5872

02:51 <gbruno> [github] kintel ready_for_review pull request #5872 (Refactoring to clean up and align CGAL geometry conversions a bit.) https://github.com/openscad/openscad/pull/5872

03:28 <gbruno> [github] kintel closed pull request #5879 (Renamed CGAL_Nef_Polyhedron to CGALNefGeometry) https://github.com/openscad/openscad/pull/5879

03:28 <gbruno> [github] kintel pushed 2 additions 28 modifications 2 removals (Renamed CGAL_Nef_Polyhedron to CGALNefGeometry (#5879)) https://github.com/openscad/openscad/commit/2f58f07831d8e6f0cf80450076a5160da2af6286

03:28 <gbruno> [github] kintel pushed 2 modifications (Make createSurfaceMeshFromPolySet() return shared_ptr to align with similar functions) https://github.com/openscad/openscad/commit/f380746349944a5a48419c99e7aecdd669133b2e

03:28 <gbruno> [github] kintel synchronize pull request #5872 (Refactoring to clean up and align CGAL geometry conversions a bit.) https://github.com/openscad/openscad/pull/5872

04:12 <gbruno> [github] kintel closed pull request #5872 (Refactoring to clean up and align CGAL geometry conversions a bit.) https://github.com/openscad/openscad/pull/5872

04:12 <gbruno> [github] kintel pushed 18 modifications 2 removals (Refactoring to clean up and align CGAL geometry conversions a bit. (#5872) * Refactoring to clean up and align geometry conversions a bit. No functional changes * Clean up some CGAL kernel naming and usage * createSurfaceMeshFromManifold() now returns a shared_ptr * Make createSurfaceMeshFromPolySet() return shared_ptr to align with similar functions) https://github.com/openscad/openscad/commit/12fe733b5a67

09:48 <cbmuser> teepee: Ping

09:48 <cbmuser> Can I send you an empty box for the keyboard? (see mail)

11:06 <gbruno> [github] damienmarchal synchronize pull request #5844 (Integration test for GUI/UX) https://github.com/openscad/openscad/pull/5844

11:09 <gbruno> [github] damienmarchal synchronize pull request #5844 (Integration test for GUI/UX) https://github.com/openscad/openscad/pull/5844

11:18 <gbruno> [github] damienmarchal synchronize pull request #5844 (Integration test for GUI/UX) https://github.com/openscad/openscad/pull/5844

12:12 <Guest6> / Maße in mm (1 m = 1000 mm)

12:12 <Guest6> / ------------------- Hauptkörper -------------------

12:12 <Guest6> module tractor_body() {

12:12 <Guest6>     color("red")

12:12 <Guest6>     translate([0, 0, 500])

12:12 <Guest6>         cube([2000, 1000, 800]); // Länge, Breite, Höhe

12:12 <Guest6> }

12:12 <Guest6> / ------------------- Kabine -------------------

12:12 <Guest6> module cabin() {

12:12 <Guest6>     color("gray")

12:12 <Guest6>     translate([1300, 200, 1300])

12:12 <Guest6>         cube([600, 600, 500]);

12:12 <Guest6> }

12:12 <Guest6> / ------------------- Vorderachse -------------------

12:12 <Guest6> module front_wheels() {

12:12 <Guest6>     translate([200, -200, 200])

12:12 <Guest6>         wheel_pair(600, 200); // Durchmesser, Breite

12:12 <Guest6> }

12:12 <Guest6>     front_wheels();

12:12 <Guest6>     rear_wheels();

12:12 <Guest6> }

12:12 <Guest6> steyr_tractor();

15:02 <gbruno> [github] kintel pushed 1 modifications (Build Qt6 with TestLib) https://github.com/openscad/openscad/commit/59aa1ebc386b3c3eb487b9b6904520046a03b675

15:21 <InPhase> kintel: I've mostly been doing a "do not use" approach to minkowski for a while, for obvious performance reasons. But I have a number of intricate slow old designs using it which I can fire up for testing when you give the word. I will do my best to break everything.

15:22 <InPhase> Also a bunch of hull stuff, but I do use that more periodically in restricted settings.

17:27 <kintel> InPhase thx. In general I think we may be able to shave some time off hull performance. Minkowski seems to mostly be slow due to CGAL's slow convex decomposition algorithm. That's fixable, but we'd have to write a new algo, which is time consuming and requires better data structures etc. - I was hoping Manifold would get there, but it's not very high priority for them. Makes sense, as minkowski is a bit of an esoteric operator outside of OpenSCAD ;)

17:32 <InPhase> kintel: Off the top of my head, and not checking the existing solution... Step 1 should probably be an iteration over points to build a list of concave out and saddlepoints, as connected points that are not concave out nor saddle points do not need to be decomposed from each other.

17:34 <kintel> I haven't researched it yet :) CGAL just feels ridiculously slow. It does feels like, what you say, a generalization of greedy 2D ear clipping might give us a lot of mileage.

17:37 <InPhase> My working hypothesis is that minimizing the amount of decomposition would have the greatest benefits, because then there's less to union again.

17:38 <InPhase> Like, do you even need convex decomposition on a sphere?

18:47 <kintel> InPhase right, so growing local convex regions.. I was thinking that perhaps decomposing into tetrahedrons would be a good first step, then re-combining them into convex blobs. I'm sure people have written papers on this ;)

18:55 <InPhase> They probably have! Who knows if any of them are comprehensible. ;)

22:38 <nomike> Hi

22:38 <nomike> Should users new tp OpenSCAD use the 2021 release or should I better recommend them the nightly version?

22:41 <InPhase> nomike: The nightly is probably the best bet right now.

22:41 <InPhase> It's in a pretty good state, and close to ready for a release.

22:42 <nomike> Ahh...cool...so I can expect a new release in the forseeable future?

22:46 <InPhase> Foreseeable, yes. I wouldn't bet on this month, as no checklist has been started for it. :)

22:46 <InPhase> (At least not that I saw yet.)

23:05 <kintel> nomike Foreseeable as in you'll likely live to see it :)

23:05 <kintel> The WIP is tracked here: https://github.com/orgs/openscad/projects/2

23:06 <kintel> ..but since we're all volunteers with wildly varying availability, things can take time. My best guess would be before the end of the year

23:48 <gbruno> [github] tristanbay opened issue #5881 (Window background color no longer matches theme) https://github.com/openscad/openscad/issues/5881

23:49 <gbruno> [github] tristanbay edited issue #5881 (Window background color no longer matches theme) https://github.com/openscad/openscad/issues/5881