00:02 <teepee> depends on what you want to try

00:03 <teepee> using the current infrastructure it might work returning an ImportNode wrapped by a ColorNode

00:03 <mmu_man> I wanted to return the OFF object, but with a ColorNode around

00:03 <mmu_man> but I'm not sure how to construct that

00:03 <teepee> so in static std::shared_ptr<AbstractNode> do_import(const ModuleInstantiation *inst, Arguments arguments, const Children& children, ImportType type)

00:14 <teepee> 2 nested calls to children.instantiate() maybe?

00:19 <gbruno> [github] DeepHorizons synchronize pull request #4986 ([WIP] Add ability to set solid name when exporting STL) https://github.com/openscad/openscad/pull/4986

00:31 <mmu_man> yeah but how do I construct the Children with those context objects and such…

00:36 <mmu_man> anyway, should sleep

02:28 <gbruno> [github] DeepHorizons synchronize pull request #4986 ([WIP] Add ability to set solid name when exporting STL) https://github.com/openscad/openscad/pull/4986

02:48 <pca006132> guso78: probably just old code

03:08 <pca006132> guso78: after looking at the code, I think it is for fixing manifoldness issue

03:09 <pca006132> or lazy, idk

03:09 <pca006132> orientToBoundAVolume is for fixing manifoldness issue iirc

03:41 <gbruno> [github] pca006132 synchronize pull request #4991 (fix manifold stuff) https://github.com/openscad/openscad/pull/4991

03:45 <kintel> pca006132 I think can just make the CGAL cache cache both nef, hybrid and manifold geometries

03:46 <pca006132> sounds good

03:46 <pca006132> so the normal cache is only for PolySet right?

03:46 <kintel> ..and just shuffle thing around to accommodate

03:46 <kintel> Still think it would be good with some typing to make it more clear what's going on

03:50 <kintel> PolySet and Polygon2d

03:50 <kintel> It all needs proper cleanup, but that's a larger challenge

07:26 <gbruno> [github] DeepHorizons synchronize pull request #4986 ([WIP] Add ability to set solid name when exporting STL) https://github.com/openscad/openscad/pull/4986

07:27 <gbruno> [github] DeepHorizons synchronize pull request #4986 ([WIP] Add ability to set solid name when exporting STL) https://github.com/openscad/openscad/pull/4986

07:29 <gbruno> [github] DeepHorizons synchronize pull request #4986 ([WIP] Add ability to set solid name when exporting STL) https://github.com/openscad/openscad/pull/4986

07:44 <guso78> Does Manifold also have properties to faces in instead of vertices?

08:11 <lockywolf> Has anyone designed clothing in OpenSCAD?

08:12 <lockywolf> Specifically, I want to design a holster for a smartphone.

09:39 <J24k5> lockywolf there are some protection sleeves/cover for smartphones

09:39 <lockywolf> J24k5: could you share a link?

09:40 <J24k5> https://www.yeggi.com/q/openscad+phone+case/

09:42 <J24k5> lockywolf or here https://cults3d.com/en/3d-model/gadget/iphone-13-pro-mockup-mechanical-dummy-model-openscad

09:43 <lockywolf> J24k5: thank you!

09:44 <J24k5> lockywolf i especially love this https://www.reddit.com/r/openscad/comments/1adcw41/i_am_completely_blind_and_this_is_my_first/  as it is done by a blind person and quite nice

09:56 <mmu_man> wow :)

11:40 <Scopeuk> https://github.com/vosen/ZLUDA well this is interesting and also disappointing it's basically been open sourced because it got defunded

11:41 <Scopeuk> I believe the blind person said they were exporting fixed viewpoints and getting chat gpt to describe the images in great detail

11:43 <Scopeuk> It is a tremendous effort, I ponder how things can be made better and all I ever manage to come up with is some sort of force feedback device or a moving bed of pins display

11:56 <pca006132> guso78: no, but I think you can do face properties by using vertex properties?

11:58 <pca006132> hmm, not vertex properties, but face id

11:59 <pca006132> https://manifoldcad.org/docs/html/structmanifold_1_1_mesh_g_l.html#ac6e7182a9cef30c9644ee167fbff30e1

12:00 <pca006132> note that this is not well tested and may have bugs

13:52 <pca006132> I tried using https://github.com/akuukka/quickhull/ instead of CGAL's hull function, as it should be faster due to floating point arithmetics

13:53 <pca006132> it is indeed faster, but it seems that the convex decomposition generated by CGAL frequently hit the "failed to solve horizon edge" condition, annoying...

13:57 <pca006132> it may be interesting to try the one in VHACD

18:33 <gbruno> [github] kintel pushed 2 modifications (Allow storing any engine-specific geometry in CGALCache) https://github.com/openscad/openscad/commit/2846a1cabfcc4520db957a16facbd2f1471226b9

18:39 <guso78k> CGALCache  appears to act like a cuckoo's  mother XD

18:44 <gbruno> [github] kintel opened pull request #4996 (Minor caching cleanup) https://github.com/openscad/openscad/pull/4996

19:06 <kintel> guso78 yeah, the way fast-csg and Manifold was integrated, they just latched onto CGAL using a bunch of ifdefs and dynamic_casts, but some stuff was forgotten..

19:06 <kintel> Needs to be separated properly, but that's larger task

19:55 <gbruno> [github] kintel synchronize pull request #4996 (Minor caching cleanup) https://github.com/openscad/openscad/pull/4996

19:55 <gbruno> [github] kintel closed pull request #4991 (fix manifold stuff) https://github.com/openscad/openscad/pull/4991

19:55 <gbruno> [github] kintel pushed 2 modifications (Allow storing any engine-specific geometry in CGALCache) https://github.com/openscad/openscad/commit/7277530aa4fbfc12b079c3cb41e12f8173be69e6

19:59 <kintel> https://github.com/openscad/openscad/pull/4987 - this PR is very odd. Feels like someone randomly copied and pasted some troubleshooting ideas without regard for what OS is being used?

20:30 <guso78k71> pca006132 I have some issues with Manifold faceIDs ...

20:30 <guso78k71> https://bpa.st/GHUA

20:32 <guso78k71> the faceID's in the output are always the  index location  the face in the Input,  I never find my supplied data back in the output

20:32 <guso78k71> is setting faceIds in meshGL the correct way to supply custom faceIds or is something else required ?

21:26 <Zr40> I just updated to the latest nightly (2024.02.15). In the F5 preview, objects used in difference() are now visible (even those parts outside of the object being differenced) and are shown "inverted" (it renders the faces from the inside, not the outside). This did not happen in 2024.02.11. The F6 render seems unaffected. Is this a known issue?

21:30 <Zr40> it looks just like F12, except with F12 it's not inverted.

21:31 <guso78k71>  tried to use runOrgialID instead but i got vec out of range.

21:31 <guso78k71> its not clear to me how to use reserveID's  , it always returns 1 ?

21:35 <Zr40> this is what it looks like with F5: https://dev.zr40.nl/openscad/20240215-f5.png

21:35 <Zr40> and with F12: https://dev.zr40.nl/openscad/20240215-f12.png