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
Sauvin has quit [Read error: Connection reset by peer]
Sauvin has joined #openscad
TheAssassin has quit [Remote host closed the connection]
TheAssassin has joined #openscad
TheAssassin has quit [Remote host closed the connection]
TheAssass1n has joined #openscad
J23 has quit [Quit: Client closed]
J23 has joined #openscad
LordOfBikes has quit [Ping timeout: 268 seconds]
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
LordOfBikes has joined #openscad
J23 has quit [Quit: Client closed]
J23 has joined #openscad
fedorafansuper has quit [Quit: Textual IRC Client: www.textualapp.com]
srk- has joined #openscad
srk has quit [Read error: Connection reset by peer]
srk- is now known as srk
kintel has joined #openscad
<kintel> If people with some Windows experience have any ideas, this really puzzles me:
<kintel> I managed to get a small example program to request an OpenGL context using my GPU over ssh. Great!
<kintel> However, when I changed OpenSCAD to use the same technique -> still no dime.
<kintel> ..so I did what every engineer sometimes does: I started deleting code from OpenSCAD until the only thing left was creating an OpenGL Context -> still nothing
<kintel> ..then I changed it even more to the point where OpenSCAD now is identical to my working test program -> nothing
<kintel> ..then I copied openscad.exe to the location of my working test program and renamed it to offscreen.exe -> now it works
<kintel> WTF?!?
<InPhase> kintel: ...
<InPhase> kintel: Registry shenanigans?
<kintel> Well, that or a kernel backdoor allowing "offscreen.exe" to do magic stuff
<InPhase> lol
<kintel> But how would that even happen? It's not like I even know how to change registry : /
<kintel> I can always reinstall Windows and try to reproduce, but I like my life too much ...
<InPhase> Well, OpenGL does look into the registry to figure out what to load. https://learn.microsoft.com/en-us/windows-hardware/drivers/display/loading-an-opengl-installable-client-driver But I can't quite figure how the program name would get into this problem.
<InPhase> To be clear, it still fails when you leave it as openscad.exe in a new blank directory?
<InPhase> kintel: Additional question... Is there an offscreen.manifest file in that directory?
<InPhase> Maybe test both openscad.exe and offscreen.exe versions of OpenSCAD in a blank directory.
<kintel> Huh, Moving files around: As long as the file is called "offscreen.exe" it works in any folder
<kintel> but renaming it to anything else fails
<InPhase> onscreen.exe fails too?
<kintel> yes
<InPhase> Did you do some installer for this offscreen.exe example you're comparing to?
<kintel> offscreen.com fails too
<kintel> No, it's just a cmake-based example program I built
<InPhase> lol
<InPhase> Hmm.
<InPhase> Was there a make install?
<InPhase> Or... anything like that?
<kintel> No, it's a slimmed-down version of this with only the win32 code: https://github.com/kintel/offscreen
<kintel> I guess Windows must have asked me once if I trust "offscreen.exe", and it has decided that the filename is good
<kintel> I understand how how writing viruses is so tempting on Windows ;)
<kintel> This is all in an msys2 shell under ssh.
<kintel> In GUI mode, everything works
<kintel> (GUI mode == RDP)
<InPhase> So, maybe also check that cmake carefully.
<InPhase> Starting maybe with a thorough scan for things like HKLM and HKEY?
<InPhase> I don't know what exactly it will look like, but the magic came from somewhere.
<InPhase> Or do a regedit and search the whole thing for offscreen.exe
<kintel> I handwrote my cmake, so it's super-minimal and doesn't have any CPACK or similar stuff
<kintel> nothing in registry either.
<kintel> Could be that the ssh server is running as a special user, which has its own settings
ur5us has quit [Ping timeout: 250 seconds]
hrberg has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
hrberg has joined #openscad
Sauvin has quit [Ping timeout: 255 seconds]
Sauvin has joined #openscad
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<InPhase> kintel: I googled for what sort of OpenGL settings have an application name in them, and found these multipipe settings mentioned. Now maybe not that, but it does have a list of config files. Maybe running under msys2 under ssh brings you access to files like a $HOME/.Xdefaults as mentioned there that would not otherwise be considered by Windows?
<InPhase> Perhaps the ssh session is setup in some manner to use these in anticipation of you doing some sort of X-forwarding.
<InPhase> Otherwise, I don't know. If not that, then more info needed I guess. :)
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
fling has quit [Remote host closed the connection]
fling has joined #openscad
guso78 has joined #openscad
<J23> InPhase .. oliver fixed ochafik.com/openscad2    now the examples are included and working
fedorafan has joined #openscad
<gbruno> [github] alidaf closed issue #4579 (Manifold: Decimates embossed text or images) https://github.com/openscad/openscad/issues/4579
gunnbr__ has joined #openscad
vsellier_ has joined #openscad
pah has joined #openscad
zauberfisch_ has joined #openscad
vsellier has quit [Killed (erbium.libera.chat (Nickname regained by services))]
vsellier_ is now known as vsellier
BZK[m] has quit [*.net *.split]
gunnbr_ has quit [*.net *.split]
pa has quit [*.net *.split]
zauberfisch has quit [*.net *.split]
castawayc has quit [*.net *.split]
HimeHaieto has quit [*.net *.split]
HimeHaieto has joined #openscad
fedorafan has quit [Ping timeout: 250 seconds]
fedorafan has joined #openscad
guso78 has quit [Quit: Client closed]
pah is now known as pa
<teepee> Olivier :)
<teepee> libraries are fine too, nice! use <ub.scad> Servo();
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
Ckat has quit [Ping timeout: 240 seconds]
Ckat has joined #openscad
srk has quit [Quit: ZNC 1.8.1 - https://znc.in]
srk has joined #openscad
castaway has joined #openscad
BZK[m] has joined #openscad
castawayc has joined #openscad
gwillen has quit [Ping timeout: 256 seconds]
gwillen has joined #openscad
mikolajw has quit [Ping timeout: 252 seconds]
guso78[m] has quit [Ping timeout: 268 seconds]
JordanBrown[m] has quit [Ping timeout: 268 seconds]
Notkea has quit [Ping timeout: 246 seconds]
Cadair has quit [Ping timeout: 264 seconds]
xloem[m] has quit [Ping timeout: 264 seconds]
evils[m] has quit [Ping timeout: 252 seconds]
agnes_de_lion[m] has quit [Ping timeout: 268 seconds]
BZK[m] has quit [Ping timeout: 248 seconds]
LinuxHackerman has quit [Ping timeout: 260 seconds]
ur5us has joined #openscad
evils[m] has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
agnes_de_lion[m] has joined #openscad
JordanBrown[m] has joined #openscad
guso78[m] has joined #openscad
evils[m] has quit [Quit: Bridge terminating on SIGTERM]
JordanBrown[m] has quit [Client Quit]
agnes_de_lion[m] has quit [Client Quit]
guso78[m] has quit [Client Quit]
Notkea has joined #openscad
mikolajw has joined #openscad
LinuxHackerman has joined #openscad
xloem[m] has joined #openscad
evils[m] has joined #openscad
Cadair has joined #openscad
BZK[m] has joined #openscad
JordanBrown[m] has joined #openscad
guso78[m] has joined #openscad
agnes_de_lion[m] has joined #openscad
fedorafansuper has joined #openscad
fedorafan has quit [Ping timeout: 246 seconds]
ur5us has quit [Ping timeout: 265 seconds]
guso78[m] has quit [Ping timeout: 240 seconds]
evils[m] has quit [Ping timeout: 240 seconds]
BZK[m] has quit [Ping timeout: 240 seconds]
mikolajw has quit [Ping timeout: 240 seconds]
LinuxHackerman has quit [Ping timeout: 240 seconds]
Cadair has quit [Ping timeout: 248 seconds]
xloem[m] has quit [Ping timeout: 248 seconds]
JordanBrown[m] has quit [Ping timeout: 265 seconds]
Notkea has quit [Ping timeout: 265 seconds]
agnes_de_lion[m] has quit [Ping timeout: 256 seconds]
LinuxHackerman has joined #openscad
BZK[m] has joined #openscad
xloem[m] has joined #openscad
evils[m] has joined #openscad
mikolajw has joined #openscad
JordanBrown[m] has joined #openscad
GNUmoon has quit [Ping timeout: 255 seconds]
GNUmoon has joined #openscad
Guest18 has joined #openscad
Cadair has joined #openscad
Guest18 has quit [Client Quit]
guso78[m] has joined #openscad
agnes_de_lion[m] has joined #openscad
Notkea has joined #openscad
fedorafansuper has quit [Ping timeout: 252 seconds]
fedorafan has joined #openscad
<pa> Hi folks, OT question: suggestion for joining a TPU part split in 2?
<buZz> interlocking connectors? :)
<buZz> here's a test of 30 different 'glues' with TPU parts ; https://www.youtube.com/watch?v=Oqh01j7ZG_s
<pa> thanks, checking
<pa> thought about interlocking, but going to be hard to design
<J23> tpu .. solder iron to fuse together
<pa> essentially i'm trying to design a magnetic cap intended to block air flow, and i'm trying to have a tpu "sealing" that is supposed to host magnets. Since the surface isn't flat but conical, the magnets sockets arent lying flat, so i have to split the sealing in 2
guerd871 has joined #openscad
guerd871 has quit [Remote host closed the connection]
<buZz> pa: yeah i guess those solvents in the vid are the solution then
<buZz> for airtightness
<buZz> magnets are that airtight btw :P
<J23> flame polishing works very well on tpu
<pa> interesting.. first time i hear about flame polishing.. i saw a video joining tpu with soldering iron but not easy in my case
<buZz> all thermoplasts can do flame polishing, i bet?
<pa> i was also thinking to remelt the top of one part somehow (like with a very hot plate) and then just press the other part over it
<J23> sure but tpu get a very liquid surface very fast
<pa> https://i.imgur.com/BMYR7Lv.png this is roughly it, the exploded view is cropped to show where the sockets are supposed to be although i was planning to slice it along xy to create printable parts
<pa> although maybe i could just leave the sockets open on the side and then put something to cork them in
<pa> maybe making a narrowing to allow pushing in but then they would stick inside
<pa> probably easier than joining 2 parts
<pa> although the problem would be bridges in tpu, i can imagine not staying straight very easily
califax has quit [Ping timeout: 255 seconds]
califax has joined #openscad
<J23> leave pockets for the magnets and then fill with liquid (hot ) tpu from the extruder
<Scopeuk> if it's just pockets you can put a pause in the gcode
<Scopeuk> have it print the pockets (with some space above) and then pause to inset the magnets, continue printing after inserted
<pa> good idea
<pa> Scopeuk: the problem with that is that the pockets aren't flat
<J23> but they are just tilted a little so if you stop at the right layer there should be just enough open to slide them in
<pa> i fear the nozzle might hit the magnets while moving
<pa> i think filling is better idea
<J23> as it is TPU with these  https://upload.wikimedia.org/wikipedia/commons/2/28/Sicherungsringzange_Welle.JPG you can stretch open to get them in
snaked has quit [Quit: Leaving]
<pa> yes, or maybe just narrow the opening a bit and push the cylindrical magnets in
<J23> if you have a solder iron with temp setting and a blade insert (like TS-100 ) you can model the part to your liking after
<J23> or leave the magnet pocket open from below  with a bit smaller opening .. so you press them in
<pa> ah good idea.. i don't have that insert but shouldnt be hard to find
<pa> i have the iron that came with an atten 8586 (i think)
<J23> a scraper or putty knife and a stove top will work with some practice
<J23> use the hot air from the atten to smooth the TPU
<J23> I am also having a bowl with cold water to stop the softening after the surface is smooth
<pa> ah good idea, thanks for the tip!
<pa> can chatgpt do openscad?
<teepee> no
<pa> useless :-D
<buZz> pa: yes
<buZz> well
<buZz> poorly
<buZz> :D
<buZz> but its getting very confused about syntax quickly
<pa> it sucks at qml too
<teepee> if you need bad useless openscad code, just copy random stuff out of stackoverlow questions ;-)
<J23> gpt4 can read in pages of instructions .. so there is a chance to get at least the syntax right
<buZz> there's some python module to autoimport&execute 'top stackoverflow entry'
<buZz> what -chatgpt- does better than gpt generally is conversating
<buZz> so changing previous answers based on future input
<buZz> and apologizing for it :P
<buZz> but gpt in general is amazing, github copilot is gpt too, i bet that can tab-complete plenty openscad?
<buZz> gee, github.dev doesnt even do syntax highlighting for .scad files? laame
banachtarski has joined #openscad
fedorafan has quit [Ping timeout: 252 seconds]
fedorafan has joined #openscad
TheCoffeMaker has quit [Ping timeout: 240 seconds]
mechadense has joined #openscad
fedorafan has quit [Ping timeout: 248 seconds]
fedorafan has joined #openscad
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
ali1234 has joined #openscad
kintel has joined #openscad
<kintel> InPhase I got one more datapoint: If I ssh into Windows _without_ any msys stuff, I get the same behavior.
<kintel> -> I ssh into the default Window shell and run the last official OpenSCAD binary: openscad.exe --info => Framebuffer Object extension not found by GLEW
<kintel> -> I rename openscad.exe to offscreen.exe: offscreen.exe --info => OpenGL Version: 4.6.0 NVIDIA 528.49
<kintel> ^ This is OpenSCAD-2021.01
<InPhase> kintel: Sometimes people ask me why I try to avoid doing Windows development.
<InPhase> kintel: I appreciate that you will not ask me this question. ;)
<Scopeuk> I know some graphics drivers try to "be clever" and mess with settings based on the exe name but this goes well beyond that
<kintel> haha, I'm really starting to regret installing Windows :/
<InPhase> kintel: But, you're doing the good community service here on trying to sort out the problems with the Windows builds.
<kintel> Problem is that as an engineer, I find this really hard to let go : (
<kintel> Perhaps I'll wait for someone else to run an ssh server under Windows and check if offscreen.exe works for them :)
<kintel> If we could get this to work, that would be a totally reasonable Windows dev. env; practically a fully transparent VS Code + Terminal, just happens to be Windows under the hood
<Scopeuk> kintel does an arbitrary third name work?
L29Ah has left #openscad [#openscad]
<Scopeuk> I'm wondering if openscad.exe has been associated with the rdp login or some similar madness
<kintel> Scopeuk No, only offscreen.exe
<kintel> even offscreen.com fails
<Scopeuk> thats very odd
<kintel> I suspect it is because I built an offscreen.exe myself as an offscreen context test, and the system must somehow have taken a liking to my program
<kintel> ..but registry claims to not know about it, and the only difference is when executing stuff in an ssh session
<kintel> (using the official Microsoft ssh server for Windows 11 setup)
<kintel> Perhaps I should reboot.
<kintel> I also didn't activate Windows yet, and it's possible that offscreen.exe existed before Windows started to complain about activation
<kintel> But yeah, caching sounds reasonable
<InPhase> kintel: You can redo your offscreen.exe build process as offscreen2.exe and see if you can trigger a second copy of a name that acts this way.
<InPhase> kintel: If so, I have a solution for how you can achieve "works for me"...
<InPhase> But that is not a satisfying solution if you're trying to make sure this works as a general solution and will be comparable to what works for others.
<InPhase> Still, triggering a second copy might confirm for you that some magic happened and that the name is not special.
<kintel> InPhase Tried that yesterday, modified cmake to build an offscreen2 target, and the offscreen2.exe does not work either
<InPhase> There is an offscreen.exe by Apple, so I'm wondering if this is some hardcoded special name.
<InPhase> Which would be absolutely terrible.
<kintel> Hardcoded special name sounds far-fetched, but can only really be verified by someone else trying to reproduce
<Scopeuk> I did try searching around didn't spot anything
<kintel> Reboot didn't change anything, I'll try buying a Windows license next
<InPhase> Yes, it's so terrible I have to hope it's wrong. But there are not many explanations for what you're experiencing.
<Scopeuk> I guess you could try "set SESSIONNAME=Console" to see if that kicks it but really shouldn't
<Scopeuk> woudlent explain the differnce by exe name either
qeed_ has quit [Quit: qeed_]
<Scopeuk> the only other thought would be that I know there are settings in the nvidia driver "per application" which I suppose could conceptually do some low level manipualation
qeed has joined #openscad
<kintel> Scopeuk right, I can try disabling the nVidia GPU and use the iGPU instead.
<kintel> SESSIONNAME did nothing
TheCoffeMaker has joined #openscad
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
L29Ah has joined #openscad
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<gbruno> [github] christophberger opened issue #4580 (Mac image appears not to be properly PGP signed) https://github.com/openscad/openscad/issues/4580
fedorafan has quit [Ping timeout: 248 seconds]
fedorafan has joined #openscad
ali1234 has quit [Ping timeout: 250 seconds]
kintel has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
<kintel> Scopeuk Good idea about nVidia driver: If I switch to my Intel iGPU, GPU access over ssh doesn't work at all
<Scopeuk> Up side: consistency, down side. Awe damn it isn't 2orking
<gbruno> [github] ochafik edited pull request #4533 (Use Manifold for much faster & multithreaded CSG & minkowski operations) https://github.com/openscad/openscad/pull/4533
fedorafan has quit [Ping timeout: 248 seconds]
fedorafan has joined #openscad
castaway has quit [Ping timeout: 268 seconds]
<J23> round to square adapter https://imgur.com/a/RLT8FwB  with extra steps
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<guso78[m]> Haha, Looks Like a xxx toy...
snaked has joined #openscad
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
fedorafan has quit [Ping timeout: 265 seconds]
fedorafan has joined #openscad
fedorafan has quit [Ping timeout: 248 seconds]
fedorafan has joined #openscad
Number3[m] has joined #openscad
TheAssass1n is now known as TheAssassin
TheAssassin is now known as TheAssass1n
TheAssass1n is now known as TheAssassin
ur5us has joined #openscad
<joseph_> teepee InPhase: Did you find out anything more yet about GSoC mentor registration? If it's gotten set up then I would indeed like to apply again this summer. My preference remains a rendering-related project, but as I mentioned before if that won't work out then I could do one of the other ideas.
<joseph_> I was thinking a good way to frame the project scope is restructuring to more consistently separate the functions that use legacy OpenGL, resulting in a clearly consolidated path that uses VBOs for rendering. Then sometime in the future, a flag could be set to deprecate the fixed-function calls. Right now those controls are split across multiple ExperimentalVxO feature checkboxes. (Out of necessity in GSoC 2022 I already made a
<joseph_> slight adjustment to when the modern methods get enabled--previously, it additionally depended on whether Show Edges was activated.) kintel do you think I should expand on these thoughts in a draft project proposal document that I'd share for potential feedback? Or do you recommend I instead just plan on submitting a non-graphics project idea?
castaway has joined #openscad
L29Ah has quit [Ping timeout: 265 seconds]
fedorafan has quit [Ping timeout: 260 seconds]
use-value has quit [Remote host closed the connection]
use-value has joined #openscad
teepee has quit [Ping timeout: 255 seconds]
fedorafan has joined #openscad
ali1234 has joined #openscad
guso78 has joined #openscad
fedorafan has quit [Ping timeout: 248 seconds]
castaway has quit [Ping timeout: 265 seconds]
guso78 has quit [Ping timeout: 260 seconds]
fedorafan has joined #openscad
L29Ah has joined #openscad
<InPhase> J23: Elder Wand Plumbing Services Incorporated strikes again.
fedorafan has quit [Ping timeout: 252 seconds]
fedorafan has joined #openscad
actiondave has quit [Read error: Connection reset by peer]
mechadense has quit [Ping timeout: 260 seconds]
kintel has joined #openscad