teepee changed the topic of #openscad to: OpenSCAD - The Programmers Solid 3D CAD Modeller | This channel is logged! | Website: http://www.openscad.org/ | FAQ: https://goo.gl/pcT7y3 | Request features / report bugs: https://goo.gl/lj0JRI | Tutorial: https://bit.ly/37P6z0B | Books: https://bit.ly/3xlLcQq | FOSDEM 2020: https://bit.ly/35xZGy6 | Logs: https://bit.ly/32MfbH5
J24k54 has quit [Quit: Client closed]
J24k54 has joined #openscad
<JordanBrown1> Why do we limit ranges to a million entries? It's a nuisance for benchmarking...
J24k58 has joined #openscad
J24k54 has quit [Ping timeout: 250 seconds]
mmu_man has quit [Ping timeout: 255 seconds]
<InPhase> JordanBrown1: Because we lack proper threading.
<InPhase> A few fixes are needed to give us some proper cancel buttonry. So some haphazard attempts were made to minimize the hole made when someone shoots their foot off.
teepee_ has joined #openscad
<InPhase> JordanBrown1: I don't think this is exactly a proper choice, but an intermediate state I suppose. And it's straightforward enough to work around with nesting.
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
yogadude has quit [Ping timeout: 260 seconds]
J24k58 has quit [Quit: Client closed]
yogadude has joined #openscad
J24k58 has joined #openscad
LordOfBikes has quit [Ping timeout: 264 seconds]
LordOfBikes has joined #openscad
<JordanBrown1> InPhase it's a nuisance to step through a >1M entry array. Not that it can't be done, just a nuisance. And for list comprehensions, which is probably the only real case, you can use C-style for.
<JordanBrown1> Oh, scratch that, got that backwards. We *don't* have C-style for in list comprehensions.
<JordanBrown1> Scratch scratching that. LCs do, normal don't. I misread the cheat sheet.
<JordanBrown1> It would be nice if it changed so that 1M was allowed and 1M+1 is disallowed.
<InPhase> Off by one on being convenient.
snaked has joined #openscad
JordanBrown1 has quit [Read error: Connection reset by peer]
Ahmed__ has joined #openscad
Ahmed__ has quit [Read error: Connection reset by peer]
Ahmed_ has joined #openscad
mmu_man has joined #openscad
zauberfisch has quit [Quit: No Ping reply in 180 seconds.]
zauberfisch has joined #openscad
Experimentalist has joined #openscad
mmu_man has quit [Ping timeout: 252 seconds]
<Experimentalist> I am trying to run the tests for the nightly build version 2024.02.16 (git 55be9cf86) on Windows 11 Version 22H2 (OS Build 22621.3155). Most tests if not all tests over 313of 1952 come back failed. Any ideas on why this is?
mmu_man has joined #openscad
greenbigfrog has quit [Ping timeout: 256 seconds]
greenbigfrog has joined #openscad
mmu_man has quit [Ping timeout: 264 seconds]
mmu_man has joined #openscad
<gbruno> [github] StarkRG opened issue #4992 (Some preferences are reset when program starts) https://github.com/openscad/openscad/issues/4992
<gbruno> [github] StarkRG edited issue #4992 (Some preferences are reset when program starts) https://github.com/openscad/openscad/issues/4992
mmu_man has quit [Ping timeout: 264 seconds]
Ahmed_ has quit [Ping timeout: 272 seconds]
mmu_man has joined #openscad
<pca006132> Experimentalist: can you show the error messages?
Ahmed_ has joined #openscad
<Experimentalist> pca006132, sorry I was away. Let me have a look as I ran again with the option '--rerun-failed --output-on-failure'
<Experimentalist> pca006132 can I upload the logs to somewhere? Sorry not used this chat before
Ahmed_ has quit [Ping timeout: 246 seconds]
Ahmed646aa has joined #openscad
Experimentalist has quit [Quit: Client closed]
Experimentalist has joined #openscad
Experimentalist has quit [Quit: Client closed]
arebil has quit [Quit: arebil]
Experimentalist has joined #openscad
<Experimentalist> Tests mainly failing from here:
<Experimentalist>  261/1586 Test #261: dumptest_allmodules .......................................... Passed 0.27 sec
<Experimentalist>           Start 262: dumptest_special-consts
<Experimentalist>  262/1586 Test #262: dumptest_special-consts ...................................... Passed 0.25 sec
<Experimentalist>           Start 263: dumptest_variable-overwrite
<Experimentalist>  263/1586 Test #263: dumptest_variable-overwrite .................................. Passed 0.26 sec
<Experimentalist>           Start 264: cgalpngtest_circle-tests
<Experimentalist>  264/1586 Test #264: cgalpngtest_circle-tests .....................................***Failed 0.39 sec
<Experimentalist>           Start 265: cgalpngtest_control-hull-dimension
<Experimentalist>  265/1586 Test #265: cgalpngtest_control-hull-dimension ...........................***Failed 0.42 sec
<Experimentalist>           Start 266: cgalpngtest_difference-2d-tests
<Experimentalist>  266/1586 Test #266: cgalpngtest_difference-2d-tests ..............................***Failed 0.49 sec
<Experimentalist>           Start 267: cgalpngtest_fill2-tests
<Experimentalist>  267/1586 Test #267: cgalpngtest_fill2-tests ......................................***Failed 0.53 sec
<guso78> Same Here. This IS because openscad fails to initialize gl for the offscreen View.
<mmu_man> terminate called after throwing an instance of 'std::bad_alloc'
<mmu_man> well mconvert needs some more work
<mmu_man> 4159439512 edges for a cube… that seems a bit large
<mmu_man> no tracing code in pmp-library :-(
<mmu_man> their code doesn't handle comments
<mmu_man> more patches to write
erectus has quit [Remote host closed the connection]
erectus has joined #openscad
fling has quit [Ping timeout: 255 seconds]
fling_ has joined #openscad
fling_ has quit [Remote host closed the connection]
fling has joined #openscad
fling has quit [Remote host closed the connection]
kintel has joined #openscad
<kintel> Experimentalist Did you build OpenSCAD yourself?
fling has joined #openscad
<kintel> One issue on Windows is that openscad.com isn't handled correctly when building manually, so you need to move winconsole/openscad.com into the same place where the .exe is.
<kintel> Not sure if that's what you experience, but I remember having to do that for every build
<InPhase> kintel: Can that be a CMakeLists.txt update?
<kintel> I think so - it's sorted on the CI, we should align it with a straight-up cmake build.
<kintel> I just lose patience really quickly when booted in Winmdoews ;)
fling has quit [Remote host closed the connection]
<pca006132> iirc my windows build on a windows vm does not require anything special
<pca006132> apart from installing a mesa driver on windows
<pca006132> not sure if that is needed for normal windows computers
Ahmed646aa has quit [Ping timeout: 272 seconds]
<mmu_man> oh, one file crashes half the time, that screams uninitialized
<Scopeuk> pca006132 shouldn't be, will only need Mesa if you have no hardware/base GPU driver opengl
<pca006132> yeah, I am expecting this
<pca006132> so no idea why initialization for offscreen view failed for guso78 and Experimentalist...
<pca006132> Experimentalist: note that you can do `ctest --rerun-failed --output-on-failure` to get the error output
<pca006132> the message you got are not really helpful in debugging
fling has joined #openscad
<kintel> also: openscad --info
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<mmu_man> so now I have many binary files to parse
<mmu_man> like I had no other stuff to do :D
<mmu_man> ugh, pmp writes whatever endian it seems, disregarding the specs
<mmu_man> and no flag to tell which it is
<pca006132> *just pretend every computer uses little endian*
<gbruno> [github] pca006132 synchronize pull request #4991 (fix manifold stuff) https://github.com/openscad/openscad/pull/4991
* mmu_man slaps pca006132 with a Bull DPX/2 300
<pca006132> iirc some ARM CPU can switch endianness as well
<pca006132> not really an ancient thing
<pca006132> ah and powerpc runs linux on big endian
<pca006132> :)
<mmu_man> yes
<mmu_man> it's just people used to only PC don't know any better so they just skip this question
<Scopeuk> on pc but not used to using standard ethernet messaging?
<Scopeuk> some of the power pc's can run mixed endian if I remember correctly, not even select at boot but swapable at runtime
<Scopeuk> why is anyones guess
<mmu_man> legacy
<mmu_man> IIRC you can even tell the MMU to auto-swap mappings, like to access I/O ranges to a PCI card
<gbruno> [github] DeepHorizons synchronize pull request #4986 ([WIP] Add ability to set solid name when exporting STL) https://github.com/openscad/openscad/pull/4986
J24k23 has joined #openscad
J24k58 has quit [Ping timeout: 250 seconds]
snaked has quit [Quit: Leaving]
teepee has quit [Ping timeout: 255 seconds]
teepee_ has joined #openscad
teepee_ is now known as teepee
<Experimentalist> Here's a sample of the logs after rerunning failed:
<Experimentalist> Start testing: Feb 17 17:54 GMT Standard Time
<Experimentalist> ----------------------------------------------------------
<Experimentalist> 53/1586 Testing: echotest_include-tests
<Experimentalist> 53/1586 Test: echotest_include-tests
<Experimentalist> Command: "C:/Program Files/Python312/python.exe" "d:/OpenSCAD-Nightly/tests/test_cmdline_tool.py" "--comparator=image_compare" "-c" "VENV_BIN_PATH-NOTFOUND/python.exe" "-s" "echo" "-t" "echotest" "-f" "include-tests" "d:/OpenSCAD-Nightly/openscad.com" "d:/OpenSCAD-Nightly/tests/data/scad/misc/include-tests.scad"
<Experimentalist> Directory: D:/OpenSCAD-Nightly/tests-build
<Experimentalist> "echotest_include-tests" start time: Feb 17 17:54 GMT Standard Time
<Experimentalist> Output:
<Experimentalist> ----------------------------------------------------------
<Experimentalist> d:/OpenSCAD-Nightly/tests/data/scad/misc/include-tests
<Experimentalist> d:/OpenSCAD-Nightly/tests/data/scad/misc include-tests
<Experimentalist> include-tests
<Experimentalist> run_test() cmdline: d:/OpenSCAD-Nightly/openscad.com d:/OpenSCAD-Nightly/tests/data/scad/misc/include-tests.scad -o D:\OpenSCAD-Nightly\tests-build\output\echotest\include-tests-actual.echo
<Experimentalist> using font directory: d:\OpenSCAD-Nightly\tests\data\ttf
<Experimentalist> stdout output:
<Experimentalist> text comparison:
<Experimentalist>  expected textfile: d:\OpenSCAD-Nightly\tests\regression\echotest\include-tests-expected.echo
<Experimentalist> 56/1586 Testing: echotest_use-tests
<Experimentalist> 56/1586 Test: echotest_use-tests
<Experimentalist> Command: "C:/Program Files/Python312/python.exe" "d:/OpenSCAD-Nightly/tests/test_cmdline_tool.py" "--comparator=image_compare" "-c" "VENV_BIN_PATH-NOTFOUND/python.exe" "-s" "echo" "-t" "echotest" "-f" "use-tests" "d:/OpenSCAD-Nightly/openscad.com" "d:/OpenSCAD-Nightly/tests/data/scad/misc/use-tests.scad"
<Experimentalist> Directory: D:/OpenSCAD-Nightly/tests-build
<Experimentalist> "echotest_use-tests" start time: Feb 17 17:54 GMT Standard Time
<Experimentalist> Output:
<Experimentalist> ----------------------------------------------------------
<Experimentalist> d:/OpenSCAD-Nightly/tests/data/scad/misc/use-tests
<Experimentalist> d:/OpenSCAD-Nightly/tests/data/scad/misc use-tests
<Experimentalist> use-tests
<Experimentalist> run_test() cmdline: d:/OpenSCAD-Nightly/openscad.com d:/OpenSCAD-Nightly/tests/data/scad/misc/use-tests.scad -o D:\OpenSCAD-Nightly\tests-build\output\echotest\use-tests-actual.echo
<Experimentalist> using font directory: d:\OpenSCAD-Nightly\tests\data\ttf
<Experimentalist> stdout output:
<Experimentalist> text comparison:
<Experimentalist>  expected textfile: d:\OpenSCAD-Nightly\tests\regression\echotest\use-tests-expected.echo
<Experimentalist>  actual textfile: D:\OpenSCAD-Nightly\tests-build\output\echotest\use-tests-actual.echo
<Experimentalist> ---
<Experimentalist> +++
<Experimentalist> "--enable=manifold"
<Experimentalist> Directory: D:/OpenSCAD-Nightly/tests-build
<Experimentalist> "echotest_include-recursive-test" start time: Feb 17 17:54 GMT Standard Time
<Experimentalist> Output:
<Experimentalist> ----------------------------------------------------------
<Experimentalist> include-recursive-test
<Experimentalist> run_test() cmdline: d:/OpenSCAD-Nightly/openscad.com d:/OpenSCAD-Nightly/tests/data/scad/misc/include-recursive-test.scad --enable=manifold -o D:\OpenSCAD-Nightly\tests-build\output\echotest\include-recursive-test-actual.echo
<Experimentalist> using font directory: d:\OpenSCAD-Nightly\tests\data\ttf
<Experimentalist> stdout output:
<Experimentalist> text comparison:
<Experimentalist>  expected textfile: d:\OpenSCAD-Nightly\tests\regression\echotest\include-recursive-test-expected.echo
<Experimentalist>  actual textfile: D:\OpenSCAD-Nightly\tests-build\output\echotest\include-recursive-test-actual.echo
<Experimentalist> ---
<Experimentalist> +++
<Experimentalist> @@ -1,12 +1,12 @@
<Experimentalist> -WARNING: Can't open include file 'include-recursive-test.scad'. in file ../tests/data/scad/misc/include-recursive-test.scad, line 1
<Experimentalist> -WARNING: Can't open include file '../misc/include-recursive-test.scad'. in file ../tests/data/scad/misc/include-recursive-test.scad, line 2
<Experimentalist> -WARNING: Can't open include file '../../scad/misc/include-recursive-test.scad'. in file ../tests/data/scad/misc/include-recursive-test.scad, line 3
<buZz> wtf
<buZz> learn2pastebin
<Experimentalist> Sorry, not used this before today, will work it out
<teepee> paste!
<othx> paste is https://www.3dcustomizer.net/paste for .scad files, https://bpa.st for text, https://pasteboard.co/ or https://imgur.com/ for images
<teepee> use one of those ^
<Experimentalist> Okay, thanks, got to go out now
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
knielsen has quit [Ping timeout: 240 seconds]
kintel has joined #openscad
knielsen has joined #openscad
<pca006132> I will suggest you to use pastebin or something
<pca006132> ah nvm
<pca006132> late to the game
<kintel> pca006132 I feel there is some misunderstanding of what "convex" means in the context of PolySets..
<pca006132> I think convex means 3D convex
<pca006132> but the way that part of the code want is just convex polygon
<pca006132> at least this is my interpretation...
<kintel> I think it really just tessellates because the is_convex() requires triangles to evaluate convexity
<pca006132> ah
<pca006132> I thought it is "We cannot render concave polygons, so tessellate any 3D PolySets"?
<kintel> (polygons are very often technically non-planar in floating point space)
<kintel> We could possibly improve is_convex() instead..
<kintel> There is a lot of old code in the codebase which we could clean up easier now that we have better tools/utilities available.
<pca006132> so what is the usage of that part of the code?
<pca006132> allows the renderer to check if the polyset is convex?
<kintel> is_convex() is used 22 different places in the codebase - I'd have to do some review to figure out why it's necessary to do here
<kintel> Part of me want to ifdef away anything related to CGAL : /
<pca006132> I think maybe it will be easier for me to split the things unrelated to cache into another PR
<kintel> yeah, that's probably a good idea.
noonien has quit [Quit: The Lounge - https://thelounge.chat]
<gbruno> [github] pca006132 opened pull request #4993 (fix manifold stuff (without geometry cache changes)) https://github.com/openscad/openscad/pull/4993
noonien has joined #openscad
<gbruno> [github] pca006132 synchronize pull request #4993 (fix manifold stuff (without geometry cache changes)) https://github.com/openscad/openscad/pull/4993
<kintel> I can't remember the history, but it looks like when converting from PolySet to Nef Polyhedron, we perform a hull if the object is convex: https://github.com/openscad/openscad/blob/55be9cf86a4059a96239e7b660481643975c8173/src/geometry/cgal/cgalutils.cc#L38
<kintel> ..which is why the convexity check needs to be good
<pca006132> but why do we want to do a hull?
<pca006132> is it just a performance optimization?
<kintel> Good question, perhaps just an optimization
<kintel> ..or to gloss over floating point issues when converting to Nef
<kintel> ..and is_approximately_convex() has a bad bug that non-convex polygons can yield a false positive even for very concave objects: https://github.com/openscad/openscad/blob/55be9cf86a4059a96239e7b660481643975c8173/src/geometry/PolySetUtils.cc#L141
<pca006132> ah
<pca006132> maybe we want to assert the polyset to only contain triangles
<pca006132> that can probably help us discover any error in the usage
<kintel> Perhaps even subtype it as a TriangleMesh, and use that in most places
<kintel> Right now we try to tessellate to triangles at strategic locations, but it's hard to track
<kintel> This could be an easy continuation of the PolySet refactoring saga :)
<pca006132> sounds good
<gbruno> [github] mmuman opened pull request #4994 (OFF import: Handle binary format) https://github.com/openscad/openscad/pull/4994
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<gbruno> [github] kintel opened issue #4995 (PolySet: Create a TriangleSet class/subtype and use that in most places) https://github.com/openscad/openscad/issues/4995
<mmu_man> where would I add a test file?
<mmu_man> tests/data/off/ ?
<gbruno> [github] pca006132 synchronize pull request #4993 (fix manifold stuff (without geometry cache changes)) https://github.com/openscad/openscad/pull/4993
<gbruno> [github] mmuman synchronize pull request #4994 (OFF import: Handle binary format) https://github.com/openscad/openscad/pull/4994
<gbruno> [github] mmuman synchronize pull request #4994 (OFF import: Handle binary format) https://github.com/openscad/openscad/pull/4994
<teepee> mmu_man: yep, that sounds good
<gbruno> [github] mmuman synchronize pull request #4994 (OFF import: Handle binary format) https://github.com/openscad/openscad/pull/4994
<mmu_man> seems STL code just assumes floats are IEEE
<InPhase> mmu_man: That's the file format.
<InPhase> mmu_man: And the format on all currently supported platforms.
yogadude has quit [Ping timeout: 255 seconds]
<gbruno> [github] mmuman synchronize pull request #4994 (OFF import: Handle binary format) https://github.com/openscad/openscad/pull/4994
<mmu_man> Well, PDP-11 wouldn't run it anyway due to endinanness :D
<InPhase> I don't remember if the loader handles endianness issues, but I remember implementing that in the binary stl save routine.
<mmu_man> PDP is neither little nor big :-)
yogadude has joined #openscad
<gbruno> [github] kintel closed pull request #4993 (fix manifold stuff (without geometry cache changes)) https://github.com/openscad/openscad/pull/4993
<gbruno> [github] kintel closed issue #4732 (Manifold - Minkowski fails on concave polyhedron with render mode (but works if you preview first)) https://github.com/openscad/openscad/issues/4732
<gbruno> [github] kintel pushed 9 modifications (fix manifold stuff (without geometry cache changes) (#4993)) https://github.com/openscad/openscad/commit/f5688998760d6b85d7b280300388448c162edc42
<InPhase> mmu_man: Well that is true, but it won't be failing due to endianness then. :)
<mmu_man> we'd probably need to strip some features to fit into the RAM…
<mmu_man> So, how would I go import facet colors for the preview…
<mmu_man> actually…
<teepee> good question, currently it's part of the AST so 1 import = 1 color max
<teepee> I assume that information will need to be pused into the geometry object
guso78k has joined #openscad
rv__ has quit [Quit: Leaving]
guso78k has quit [Quit: Client closed]
<mmu_man> hmm, you can't create a geometry with nodes inside?
<mmu_man> I'm still a bit lost in there
<teepee> you can, right now the imports only create a single node, but that's just how it is right now
<teepee> question is how to get those partial meshes then, you can't just pick some faces, each node needs to be a solid
<teepee> well, not for preview but we still want to do F6 I suppose
<mmu_man> ok, for OFF that won't work
<mmu_man> but…
<teepee> manifold has vertex properties which sound like the way to go, but those probably would need to be handled in PolySet too
<mmu_man> it seems the KiCad VRML export actually creates a whole tree of objects, not just a single blog of facets
<mmu_man> so probably this could be doable as is
<mmu_man> just don't know how to instanciate these nodes
<mmu_man> and the syntax looks crazy
<teepee> if the parts are solids, maybe
<teepee> do you have a sample file?
<mmu_man> too large for pastebin :D
<teepee> gist maybe?
<mmu_man> random component tests
<mmu_man> I can see color, transform and shape nodes there
<teepee> I have not installed kicad on the new notebook yet
Non-ICE has quit [Quit: Screw you guys, im going home!]
Non-ICE has joined #openscad
Non-BEAST has joined #openscad
guso78k has joined #openscad
<guso78k> in manifoldutils.cc: createMutableManifoldFromPolySet: why do we convert to a CGAL Surface mesh first ? Isn't Manifold supposed to replace CGAL ?
stefanct has quit [Excess Flood]
stefanct has joined #openscad
<mmu_man> hmm trying to apply color to the Polyset before returning for testing… doesn't work
<mmu_man> not sure if Geometry::Geometries is what I want