<sparsh-n>
Seems like having my executable built in build/ directory is all that was needed. I guess in this case, whenever someone wants to build the initial cmake call where I generate the build system (sln, vcxproj, etc) should have -DCMAKE_RUNTIME_OUTPUT_DIRECTORY_DEBUG=PathTo/build. I
<sparsh-n>
For some reason running ctest still runs into an issue. https://bpa.st/EKMA I think this is because its building the executable as openscadd.exe, although I'm not sure as to why it would append a d. I'll look into it further tomorrow.
J25k22 has quit [Quit: Client closed]
sparsh-n has quit [Quit: Client closed]
J25k22 has joined #openscad
<teepee>
sparsh-n kintel: I suppose adding one more relative path will not make things much worse but would support the native window dev layout with the debug/release subfolders
<teepee>
I think that's better than messing with the defaults overwriting them with extra cmake-variables
<teepee>
maybe a separate way of specifying a different case would be the option for really non-standard layouts where there's no way to even guess the path like on guix
Guest16 has joined #openscad
<teepee>
even then we probably need at least one of the relative options for when the binary is running in the "after-installation" layout
teepee_ has joined #openscad
teepee has quit [Ping timeout: 244 seconds]
teepee_ is now known as teepee
hyperair has joined #openscad
howiemnt has joined #openscad
<Guest16>
how unstable is the nixos package "openscad-unstable" - any experiences?
<Guest16>
(openscad-unstable seems to be a few days behind the main branch on github, hence i am asking if somebody actively curates it, or if it just tracks master)
<teepee>
I have no idea who is maintaining that
hyperair has quit [Ping timeout: 265 seconds]
califax has quit [Remote host closed the connection]
<teepee>
ah, pca006132 is here in the channel, so there might be an official answer eventually
califax has joined #openscad
<pca006132>
it tracks master, but it is updated only when we remember to do so :P
<pca006132>
because it requires some work to update
<pca006132>
and review etc.
<pca006132>
the name is just convention in nixpkgs, the most appropriate name will be openscad-nightly, I think it is usually quite stable
<Guest16>
ah, thanks for the info pca006132 - so it might be better to build from git master when using on nixos then? i guess there is no general tag for a "relatively" stable status in master, no?
<teepee>
so same as I have to do with Snap and Flatpak, it also requires manual intervention :/
<pca006132>
there is no such tag
<Guest16>
how much manual intervention is needed? is this documented somewhere?
<pca006132>
well openscad requires some time to build, so if you can use the one on nixpkgs it will be faster
<pca006132>
for nixpkgs, just edit some files, open a PR, wait for review
<pca006132>
not really a lot of work, but definitely not as simple as automated deployment
hyperair has joined #openscad
J25k22 has quit [Quit: Client closed]
J25k22 has joined #openscad
Guest16 has quit [Quit: Ping timeout (120 seconds)]
hyperair has quit [Ping timeout: 248 seconds]
bitbasher has joined #openscad
bitbasher has quit [Changing host]
bitbasher has joined #openscad
J25k22 has quit [Quit: Client closed]
J25k22 has joined #openscad
howiemnt has quit [Remote host closed the connection]
howiemnt has joined #openscad
califax has quit [Remote host closed the connection]
califax has joined #openscad
califax has quit [Remote host closed the connection]
califax has joined #openscad
bitbasher has quit [Ping timeout: 252 seconds]
califax has quit [Remote host closed the connection]
califax has joined #openscad
fling has quit [Ping timeout: 244 seconds]
fling has joined #openscad
snaked has quit [Remote host closed the connection]
J25k22 has quit [Quit: Client closed]
J25k22 has joined #openscad
L29Ah has left #openscad [#openscad]
L29Ah has joined #openscad
L29Ah has left #openscad [#openscad]
J25k22 has quit [Quit: Client closed]
J25k has joined #openscad
califax has quit [Ping timeout: 244 seconds]
hyperair has joined #openscad
bitbasher has joined #openscad
sculptor has joined #openscad
sculptor has quit [Changing host]
sculptor has joined #openscad
peepsalot has quit [Remote host closed the connection]
peepsalot has joined #openscad
L29Ah has joined #openscad
kintel has joined #openscad
<kintel>
sparsh-n: I assume the added "d" is for "debug" ?
<kintel>
Also keep in mind that on Windows, openscad.com is used for cmd-line execution and openscad.exe is used for GUI. This is due to how Windows deals with attaching stdio streams to the console
kintel has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<rapha>
InPhase: I see. So there is some experience, but also some logic there. That kind of gives me hope :)
<rapha>
And I love printf debugging ... never really warmed up to debuggers.
<InPhase>
I can use debuggers, but print debugging is just far more versatile and adaptable to different situations. I prefer to build my mindsets around print debugging. And in OpenSCAD we have no debugger. :)
<InPhase>
At scale for larger programs, I try to use log debugging, with the same mindset as print debugging.
J25k3 has joined #openscad
J25k has quit [Ping timeout: 272 seconds]
<pca006132>
The problem with debugger is in part because how bad debug informations are...
<pca006132>
oftentimes temporary variables are just optimized away, and the debugger has no way to recover that information (even in theory it should be able to)
<InPhase>
And then heisenbugs if you try to switch out of release to make things more debugger friendly. :)
howiemnt1 has joined #openscad
howiemnt has quit [Ping timeout: 276 seconds]
<sculptor>
that threading library - does it render standard threads properly
<sculptor>
i prolly need it for m4 m5 m6 screws. whatever is possible with 0.4mm nozzle and 0.1mm layer height
<sculptor>
tho 0.2mm nozzle is in mail
<InPhase>
Looks like it might be. It appears attention was put into a bunch of standards, so hopefully the author if it did testing.
califax has joined #openscad
<InPhase>
I did a lot of comparative testing when I created my threading library: https://github.com/rcolyer/threads-scad But it only covers standard metric threads for intercompatibility. I never made any attempt to support any of the other standards.
<InPhase>
m4, m5, and m6 were well tested though.
<sculptor>
impressive
<InPhase>
I was covering a gap in what was available at the time when I created that. It's possible some of the ones out since then have done a better job, since I largely moved on to other topics. But mine has kept working for me for the intervening 9 years.
TheAssassin has quit [Remote host closed the connection]
TheAssassin has joined #openscad
<InPhase>
There was an open opportunity, for example, in supporting more cross-section profiles.
<InPhase>
This is also more viable now to pursue with manifold.
<InPhase>
A big focus of my library design was efficiency, since in those days a few threads could kill your design's render time if you weren't careful with how you did it.
<sculptor>
your other oscad repos are interesting as well
<InPhase>
Some code cleanliness aspects were compromised in order to favor that efficiency, so it's not so easy to expand the functionality.
<InPhase>
Hence I both recommend its utility, but in fairness one should leave a door open to other libraries that might better fit the moment in time. :)
<sculptor>
i printed some designs with threads and results were bad
<InPhase>
Well, give mine a test. It usually "just works" if you're aiming for a standard match and your printer is in good shape.
<InPhase>
The threads and holes were designed to both match each other, and metal parts, with all the little clearance adjustments needed to make that work.
<sculptor>
i printed a camera holder with 2 screws. it's designed awfully tho. camera is in mail. when it arrives i will make mine in oscad
<sculptor>
there's a lot of free models that plain suck
<InPhase>
Yes there are. :)
<InPhase>
Sometimes people share before even trying things. And some ideas are bad. It happens.
<sculptor>
i do not intend to tweak someone's .stl
<sculptor>
gimme .scad or .step or whatever it is with primitives
<InPhase>
I've done that a few times. It's usually faster and better to start from scratch.
<sculptor>
yep
<sculptor>
did i share pic of my toilet flusher actuator openscad design. red pla in the pic that follows
<sculptor>
n20 wateproofed mini-motor, a nylon 10T rack and 9T D gear. did not print those
<sculptor>
s/gear/pinion/
<sculptor>
next design is a linear actuator petg holder and tpu bellow boot for 25mm extension - it is supposed to tip the self-closing toilet lid. and that will complete toilet designs for my home automation setup
<sculptor>
for toilet seat status i will use floor mounted 1deg lidar in a standard smoke detector case
<sculptor>
s/floor/ceiling/
<sculptor>
chatgpt made a good bellow boot parametric module in several iterations
aiyion2 has quit [Ping timeout: 244 seconds]
TheAssassin has quit [Ping timeout: 244 seconds]
teepee has quit [Ping timeout: 244 seconds]
teepee has joined #openscad
TheAssassin has joined #openscad
aiyion2 has joined #openscad
<InPhase>
sculptor: This is a tremendous effort for a simple problem. :) But I can respect that.
<sculptor>
i prefer: unique
<InPhase>
After all, why should you have to flush the toilet manually like a Neanderthal.
<sculptor>
this is optional - with a wall switch or a phone ap. regular flush button works
<sculptor>
imagine pranking your guests while using your toilet
<InPhase>
phone app. lol
<sculptor>
there's a practical usage, tho
<sculptor>
for automatic closing of the toilet lid, if left open
<InPhase>
You have kids?
<sculptor>
there was at least one case of a rat entering a flat in my building through sewer and toilet seat
<InPhase>
That's a little wild, but okay. :)
<sculptor>
lid is self-closing. you only need to nudge it a bit
<teepee>
just in case you have not used it, you can put the cursor behind a number and use ALT+Cursor Up/Down to change the number
<teepee>
I'm using that for testing the parametric stuff
<InPhase>
teepee: It vaporized some time ago.
<InPhase>
buZz: But... Can I 3D print a case for my raspi+lcd based 3D printer simulator?
<InPhase>
This needs more meta.
<fsphil>
teepee: oooh that looks spot on
<InPhase>
teepee: I would have done that with: $fa=1; $fs=0.2; offset(2) for (a=[0:120:359]) rotate(a) projection() rotate_extrude(angle=70) translate([30,0]) square(eps);
<InPhase>
For the same eps=0.01
<InPhase>
I like the projection() rotate_extrude pattern for 2D arc patterns.
<fsphil>
I wouldn't have thought to make arcs like that, nice
<buZz>
InPhase: i hope they'll add maintainance aswell :P lubrication, belt replacements, unclogging, etc
<InPhase>
fsphil: I think I learned it here, but I forget who did it first.
<InPhase>
It was quite a while ago
<teepee>
ah, nice, projection is a neat solution
<buZz>
teepee: is there a kbd shortcut to switch tabs too?
<teepee>
ctrl+tab
<buZz>
and ctrl+shift+tab for back?
<teepee>
of course
<buZz>
nice, i'll try to keep that in my brain
<buZz>
i keep hitting ctrl+pgup/pgdn
<teepee>
just mix up openscad with firefox :)
<buZz>
too many programs with tabs use that :)
<InPhase>
I don't think I've ever seen ctrl+pgup/dn for tab switching.
<teepee>
the keyboard config PR is probably ready in a decade or two ;-)
<InPhase>
But I probably wouldn't have thought to try it.
<buZz>
:)
<buZz>
teepee: sweet, i'll celebrate my 66th birthday on the release
<buZz>
:)
<teepee>
but no military parades, not even 3d printed ones :P
<buZz>
InPhase: from what i have open now, it works in ; gnome-terminal, telegram, chromium
<InPhase>
Yeah, it's 3 keys for me for the ctrl-pgup/dn too, but it does in fact work. I see it is the keyboard shortcut for xfce4-terminal, but I never noticed because I work very hard to not use tabs with that. And really with most programs. I hate tabs. I have a good window manager arrangement, and tabs just try to reinvent this poorly.
<teepee>
see that S1, S2, S3, S4? all different shift keys :D
<buZz>
heh , i have a U880 somewhere in storage
<buZz>
for some future 'breadboard z80clone'
<InPhase>
The fact that pgup/dn and home/end take 2 keys on my Framework 13 is my only real complaint about the Framework 13. Otherwise a great system. I knew the 2 keys thing at purchase, so it's no shock. Just an adaptation and less desireable trait that I decided to live with in trade for other benefits.
<teepee>
yep, that thing is Z80, so you can clone that directly
<buZz>
teepee: but that feels like utterhell, the keyboard :)
<teepee>
it probably was. this one was only *almost* my first computer
<buZz>
luckily nowadays you can cheaply buy blackberry style keyboards for diy projects
<teepee>
the 1 year waiting time saved me :D
<buZz>
iirc those are even i2c
<buZz>
teepee: we share similar age i guess :P
<teepee>
the arturo ones?
<buZz>
my first was a Sharp MZ-721
<buZz>
in ~1984
<teepee>
yes, I believe we do, just me in east germany with very limited choice
<buZz>
hmhm, behind the curtain
aiyion2 has quit [Remote host closed the connection]
<InPhase>
Wow that's a terrible looking Z1013 keyboard. What a historical blunder. lol
<InPhase>
How very East German of them, giving you an approved set of graphics elements that you may use. ;)
<teepee>
indeed
<teepee>
probably copied from some western computer though. the software copyright law was, there is no law :)
guso78k has joined #openscad
guso78k has quit [Client Quit]
aiyion2 has joined #openscad
mmu_man has joined #openscad
FluxLalonde has joined #openscad
FluxLalonde has quit [Changing host]
FluxLalonde has joined #openscad
FluxLalonde has quit [Quit: Client closed]
sparsh-n has joined #openscad
teepee_ has joined #openscad
<sparsh-n>
kintel I renamed the openscadd.exe to openscad.exe, and that fixed any issues with running openscad.com. I agree with teepee that it might be better to add a relative path to the PlatformUtils as opposed to having to specify building in the build folder. Considering there's the option to build in both Debug and Release, I'm guessing a build/Debug
<sparsh-n>
and build/Release would be more clear than having all the executables in just build. For the time being I'll try getting a non headless debug build working, as headless seems to work without any issue.
mmu_man has quit [Ping timeout: 276 seconds]
<sparsh-n>
Also is ctest supposed to run on windows builds too? I built it with MSYS2 just to see if adding the relative path would make anything slower, but when I run ctest it finds the test but basically all of them fail. Which is odd, considering it builds fine and the executable works as usual.