00:48 <kintel> peepsalot Did you figure out the GL stuff?

00:48 <peepsalot> yeah, i just wasn't using builtin opencsg

00:48 <kintel> I believe Windows still uses the old stuff. Once that's fixed, we can probably wipe the old folder. It really just needs some testing, and my Windows box didn't survive long enough for me to test it

00:49 <kintel> I ended up in a situation where i had to rename openscad.exe to "offscreen.exe" for it to be able to use the GPU. Too weird to think about

00:50 <InPhase> lol. I remember your openscad rename issue. You never figured out the source of that?

00:54 <kintel> no, but I'm convinced that the remedy is reinstall windows :)

01:16 <peepsalot> i am actually surprised and confused that clang-tidy seems to pick up issues in platform specific files that aren't being used (like PlatformUtils-win.cc when building on linux), i thought it pulled from the cmake build commands that should not be including them

02:05 <kintel> peepsalot huh, that's odd. I haven't really looked into the details, but my impression was that it would pick up compile_commands.json

02:57 <peepsalot> kintel: clang-tidy found an unused variable in some 12yr old code here https://github.com/openscad/openscad/blob/335ed8eb267e0d3436124e69e5fd5d584d3be02e/src/utils/boost-utils.cc#L88

02:57 <peepsalot> i'm just a little hesitant to simply delete because it feels like its a case that was meant to be checked for and then wasn't?

02:57 <peepsalot> but the code is hard for me to follow what exactly is going on

02:58 <peepsalot> also there is a comment link to boost svn trac issue which seems defunct

03:01 <peepsalot> basically my question would be if the input somehow already contains ".." is that being handled properly?

03:07 <peepsalot> but i guess if it hasn't caused noticeable problems in all this time its probably fine...

03:07 <kintel> I think this code was pretty much copied from that defunct discussion link

03:09 <kintel> Whenever we find thing like this, I'm considering whether we should start writing unit tests for stable interfaces

03:20 <peepsalot> i think maybe this same functionality is covered by "relative" which didn't exist way back when? https://www.boost.org/doc/libs/1_86_0/libs/filesystem/doc/reference.html#relative

03:25 <peepsalot> looks like it was added in boost 1.60, and our current dependencies list calls for 1.56+

03:28 <peepsalot> https://www.boost.org/users/history/version_1_60_0.html

03:28 <peepsalot> Added functions lexically_normal, lexically_relative, relative, and weakly_canonical. Many thanks to Jamie Allsop for his help and perseverance. Resolves tickets #1976, #5897, #6249.

03:28 <peepsalot> there is mention of the same ticket #1976

03:38 <gbruno> [github] thehans pushed 3 modifications (Fixes for clang-tidy: bugprone-unused-local-non-trivial-variable) https://github.com/openscad/openscad/commit/56e34475d8c508f727158ea2e13dec04fc28d9da

03:38 <gbruno> [github] thehans opened pull request #5379 (Various fixes for clang-tidy checks with few occurences) https://github.com/openscad/openscad/pull/5379

03:39 <kintel> peepsalot interesting.. would be nice to cut out any complex special code :)

03:41 <kintel> another question is when we can start using std::filesystem instead of boost

03:44 <nmr_scientist> what's wrong with boost?

03:52 <kintel> Nothing wrong really, but I'd prefer standard C++ - one less external dependency to worry about

04:02 <gbruno> [github] thehans pushed 3 modifications (Fixes for clang-tidy: bugprone-unused-local-non-trivial-variable) https://github.com/openscad/openscad/commit/afa677699b3d09b74b68565764ef6c5566e274e0

04:02 <gbruno> [github] thehans synchronize pull request #5379 (Various fixes for clang-tidy checks with few occurences) https://github.com/openscad/openscad/pull/5379

05:10 <pca006133> yeah, and boost is just so large that it makes compilation time longer than required

06:22 <J24k4> f0lkert the community (reddit) very much appreciate your addition https://www.reddit.com/r/openscad/comments/1g4fa6v/povray_with_openscad

06:33 <UltimateCodeWarr> That PovRay shading is ultra cool, very hard to distinguish reality from rendering.

06:35 <f0lkert> J24k4: that's very cool!

07:00 <UltimateCodeWarr> https://www.pinterest.com/pin/411657222207213284/

07:01 <UltimateCodeWarr> Had to to look up that design, 6-axis hexapod?

07:10 <peepsalot> UltimateCodeWarr: https://en.wikipedia.org/wiki/Stewart_platform

07:37 <Scopeuk> and it's inverse https://github.com/pepijndevos/spacefox (I have my own corrupted version with no buttons and a fully printed base instead of a pcb

07:37 <Scopeuk> not really used it much

07:41 <J24k4> That is weird seems the manifold repair can cause flipped normals -  the object is created with rotate_extrude and offset(±) is used - depending on the values (2 or 1.85) i get flipped normals https://imgur.com/a/wzKuiqM

07:45 <J24k4> FYI kintel

13:24 <pca006133> asking again: any windows users here? can someone help me try https://github.com/elalish/manifold/pull/984#issuecomment-2414253437 ?

14:56 <J24k4> i have windows 10 - need advice what to do

15:03 <pca006133> do you have msvc installed?

15:10 <J24k4> öh don't think so - not consiously

15:11 <J24k4> a visual studio - i had that ages ago - but i think i removed it after it started to ask for logins or other things

15:14 <pca006133> I see

15:14 <pca006133> oh msys2 is fine as well

15:15 <pca006133> I think you just need cmake and gcc installed with msys2, and you can compile manifold

15:15 <pca006133> cmake . -DCMAKE_BUILD_TYPE=Release -DBUILD_SHARED_LIBS=OFF -DMANIFOLD_DEBUG=OFF -DMANIFOLD_PAR=ON -B build

15:16 <pca006133> and then cd into the build directory and run make -j<n> (where n is the number of cores, e.g. -j4)

15:20 <J24k4> i think it was installed with unity but some licenses changed

15:20 <J24k4> so i install this? https://www.msys2.org/

15:21 <pca006133> yes

15:24 <J24k4> installation failed "something evalating rights …"

15:25 <J24k4> error while elevating access rights

15:25 <J24k4> aborted .. sorry

15:40 <pca006133> hmmm... I thought msys2 is simple to install

15:41 <pca006133> anyway, it is not very important, we just want to get rid of some optimization that we think are no longer needed, and that only affects windows users iirc

15:42 <pca006133> I added that some time ago when I did some profiling on a Windows vm and found some anomalies

21:39 <Scopeuk> It looks like that pr got merged, should be able to just use the nightly builds