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
teepee_ has joined #openscad
teepee has quit [Ping timeout: 252 seconds]
teepee_ is now known as teepee
hrberg has quit [Ping timeout: 255 seconds]
hrberg has joined #openscad
zauberfisch has quit [Read error: Connection reset by peer]
zauberfisch has joined #openscad
L29Ah has joined #openscad
LordOfBikes has quit [Ping timeout: 260 seconds]
LordOfBikes has joined #openscad
J23k97 has joined #openscad
J23k16 has quit [Ping timeout: 245 seconds]
mmu_man has quit [Ping timeout: 240 seconds]
califax has quit [Remote host closed the connection]
califax has joined #openscad
pbsds has quit [Ping timeout: 240 seconds]
pbsds has joined #openscad
teepee_ has joined #openscad
teepee has quit [Quit: bye...]
teepee_ is now known as teepee
_whitelogger has joined #openscad
qeed__ has joined #openscad
qeed_ has quit [Ping timeout: 264 seconds]
invultri has joined #openscad
<invultri> so... having designed my shed+greenhouse (basically 2x4 framing with triangle based arched roof) and now needing to make the 2d drawings with measurements and details... is this where I should just go to a regular 2d cad program (librecad or something) or is there something pragmatic I can do in openscad
misterfish has joined #openscad
<J23k97> you can do 2D and you find libraries with arrows and measurements ..  so you can use the variables you already have.  Or also using projections to get that 2D image of your object
<invultri> that is true, and I have the points, however it is the rotations that screw me over here.
<invultri> so I probably have to manually construct each view
J23k97 has quit [Quit: Client closed]
J23k97 has joined #openscad
mmu_man has joined #openscad
L29Ah has left #openscad [#openscad]
teepee_ has joined #openscad
teepee has quit [Ping timeout: 252 seconds]
teepee_ is now known as teepee
fancsali has joined #openscad
<fancsali> Hi all!
<fancsali> I've bumped into the age old issue of not being able to build on ARM32 against GLES. And then ended up on the GitHub page describing the proposed rendering improvements: https://github.com/openscad/openscad/wiki/Project:-Improve-OpenGL-rendering...
<fancsali> ... and eventually ended up on the issue specific to GLES: https://github.com/openscad/openscad/issues/292.
<fancsali> Which - having been closed - suggests this has been completely abandoned... Is there any chance, this will ever happen?
<teepee> there's currently some huge progress on opengl side
<teepee> but, it's still a spare time project so no way of giving promises
<fancsali> Of course, I do get the concept of OSS. But that's good news.
<teepee> if you *really* want to get OpenSCAD running on ARM32 the safe bet is recompiling Qt
<fancsali> When you say OpenGL side, you mean proper GL or GLES included?
<fancsali> Well, probably I can adjust the compile flags on my Qt APKBUILD, tbh. What do I need to adjust?
<fancsali> Do you happen to have any pointers?
<teepee> there's a swith for -gles and -opengl IIRC
<teepee> buZz?
<fancsali> So does OpenSCAD has some direct dependencies in proper OpenGL/GLEW/whatever, or is the issue caused indirectly: OpenSCAD depending on Qt, which in turn depends on those "desktop" things.
<teepee> -opengl desktop instead of -opengl gles2/3 if my duckduckgo is good today
<teepee> both
<teepee> but latest dev version switched to GLAD which for example now allows headless EGL on Linux, no need to start an X server anymore
<fancsali> If I try to build Qt5 with -opengl desktop on an ARM32 with only GLES, that will not fly, I recon...
<teepee> all that work combined with the effort to finally toss out all the ancient OpenGL code from 1.3 *may* allow GLES at some point
<teepee> if that's going to be GLES2 I don't know, bigger chances for GLES3 maybe
<fancsali> Interesting. So I have better chances with the dev branch...
<fancsali> ... or try to build "headless". Is that an option?
<teepee> I think Qt will fall back to software render, I had version rsunning on 32-bit raspberry pi and buZz here did self compile for some other, likely only GLES2 capable 32-bit system
<teepee> depends on your usecase
<teepee> there's also a WASM version running in local browser only
<fancsali> Yeah, tried the WASM version, rather unstable on Firefox... :(
<fancsali> Did any of you guys document that effort somewhere - I mean the RPi stuff...
<teepee> I lost my stuff and gave up eventually, taking too much time, so I only do arm64 now
<fancsali> :)
<fancsali> Well, that's my dilemma to - keep the current ARM32 convertible tablet, or buy another one. But if I upgrade, then I'd probably go for x86, and avoid the whole problem completely...
<teepee> I like a lot if it's posssible to keep things running, but sometimes that can eat quite a lot of time
<teepee> maybe buZz still has some notes, if it's not urgent hang around and see if you can catch him
<teepee> I would expect it's mostly the Qt5 compile which is huge on a slow system
<fancsali> I feel your sentiment, mate! It's such a sense of achievement and geeky kick to keep my old stuff going...
<teepee> my goal would have been in addition to automate the build so we can share the binaries on the OpenSCAD website and that was a bit additional problem
<fancsali> I'll stay around for sure...
<teepee> s/bit/big/
<fancsali> I might be able to help with automation...
<fancsali> I am wondering, however, if I go desktop Qt5 they everything will be software, or just those things, that are in GL but not in GLES?
<fancsali> I've seen some GL->GLES shims around somewhere: one would compile against desktop and then LD_PRELOAD the shim to fix the missing stuff... not sure, if that route would work.
<fancsali> Found it: gl4es, that was the name!
<invultri> nice I made openscad ctd
<fancsali> ctd?
<invultri> crash to desktop
kintel has joined #openscad
<invultri> [for (i=len(path)-1,t=[things];i>=0;t=[each path[i],t]) t] it did not like this
<invultri> ah a i=i-1 helps
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
fancsali has quit [Quit: Client closed]
kintel has joined #openscad
<kintel> fancsali I'm doing a bunch of work on the GL backend when time permits. We also had some google summer of code work done on the same topic.
<kintel> I'm almost done with my GL context rewrite, which should enable asking for a GLES or compatible context for offscreen purposes
<kintel> The next obstacle is to rewrite fixed function GL code with modern GL.
<kintel> A translating wrapper could be interesting. Perhaps not as the end result, but it could help us move gradually towards legacy free GL
<kintel> ..but as always, the rabbit hole is pretty deep...
<kintel> ..plus it tends to fork into multiple separate rabbit holes : /
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
qeed__ has quit [Quit: qeed__]
qeed has joined #openscad
<teepee> invultri: endless loops indeed tend to break things ;-)
<teepee> which is actually good so they are not endless after all :D
cart_ has joined #openscad
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
pbsds has quit [Ping timeout: 240 seconds]
cart_ has quit [Ping timeout: 240 seconds]
pbsds has joined #openscad
<gbruno> [github] chotchki closed issue #4649 (Manifold crashes on second render of the SAME model on macOS) https://github.com/openscad/openscad/issues/4649
<gbruno> [github] t-paul closed issue #4685 (Manifold: Hit F6 twice causes assertion failure) https://github.com/openscad/openscad/issues/4685
<invultri> teepee: of course, but a crash to desktop is kinda unexpected
<teepee> true, but oom-killer logic still seems shaky even though it's quite a bit better than years ago
<teepee> and builtin checks in OpenSCAD are only catching some of the geometry runaways, for the language evaluation it's not really working
mmu_man has quit [Ping timeout: 252 seconds]
aiyion has quit [Remote host closed the connection]
aiyion has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 252 seconds]
teepee_ is now known as teepee
cart_ has joined #openscad
L29Ah has joined #openscad
<invultri> hmm USE does not give me access to constants in the imported file?
L29Ah has quit [Ping timeout: 244 seconds]
L29Ah has joined #openscad
L29Ah has left #openscad [#openscad]
mmu_man has joined #openscad
<teepee> invultri: correct, only functions and modules
<invultri> so import it is I guess
<invultri> include*
<teepee> yeah, either split the constants and do import of those, or if it's not too much stuff, functions might work too
LordOfBikes has quit [Remote host closed the connection]
LordOfBikes has joined #openscad
<InPhase> invultri: Or functions as constants.
<InPhase> Oh, yeah, what teepee said.
cart_ has quit [Ping timeout: 252 seconds]
<invultri> openscad-nightly bostuin11/greenhouse3.scad -o test.png --autocenter --viewall --colorscheme=BeforeDawn --camera="0,0,0,0,0,-90" --imgsize="2970,2100" <-- is there a way I can 'zoom in' ?
<InPhase> invultri: Well take off viewall and then use the camera arguments to set the camera distance?
<InPhase> You look like you're trying to use 6 of the 7 version camera arguments, but then that will be interpreted as the 6 version, but you probably did not mean z=90. So read that --help line more carefully for camera.
<InPhase> You can use either version, but you can't get the 7 one when you give 6 values.
<invultri> I think I did mean that z value, since I want a top down view
<invultri> but in the 7par top down seems to be the default
<invultri> openscad-nightly bostuin11/greenhouse3.scad -o test.png --colorscheme=BeforeDawn --camera="0,0,0,0,0,0,3200" --imgsize="2970,2100" <-- this did the trick, not sure 'why' 3200 but it seems to be about right
qeed_ has joined #openscad
qeed has quit [Ping timeout: 258 seconds]
<teepee> heh, successfully soldered 6 ws2812-2020 leds, 2mm by 2mm is not that big :)
cart_ has joined #openscad
<J23k97> nice i have done 3535 which was fun already ..
<J23k97> but yours are just a ¼
<J23k97> 5¢ per LED .. they got cheap
<InPhase> Yeah. That kind of stuff can be very tough... Except I can't for the life of me figure out what we're all doing wrong. Just watch here: https://youtu.be/HF0E8WeEUVM?t=1376
<othx> InPhase linked to YouTube video "Inside a Small Chinese Electronics Factory - From the Archives" => 1 IRC mentions
<teepee> I was lazy and ordered a bunch on amazon, still only 8¢
<teepee> that macro keyboard kindly came with the pads already :)
<teepee> oh, is that the new strange parts video?
<teepee> I saw a couple of days ago that he's back making videos, that's awesome. that recovery must have been maybe 2 years or so=
<gbruno> [github] jbinvnt edited pull request #4752 (GSoC 2023: Final Checkpoint) https://github.com/openscad/openscad/pull/4752
cart_ has quit [Ping timeout: 246 seconds]
invultri has left #openscad [#openscad]
<joseph_> teepee: I just submitted the form for my final GSoC evaluation. The submission links to #4752 which contains my report, but I'm expecting that PR to get closed and serve just as a historical record like last year. #4738 is the one that actually contains my work on the macro.
L29Ah has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 252 seconds]
teepee_ is now known as teepee
redlizard has quit [Ping timeout: 240 seconds]
redlizard has joined #openscad
<buZz> i dont think i stored any notes about my arm32 build, it was pretty straightforward, after installing mesa at least :)
qeed has joined #openscad
qeed_ has quit [Ping timeout: 260 seconds]