<InPhase>
guso7810: It does not require the python to be in the path. It only requires cmake's find routine to be able to locate the installed python.
<InPhase>
guso7810: From there everything is by full path.
<InPhase>
guso7810: If you're using my oscadcleanbuild such that your build is in the "build" directory, then this will run the tests: https://bpa.st/5SMSS
<InPhase>
guso7810: Oh, and since testing just changed significantly, you should definitely merge in updates to master to the python-support branch before making any edits.
<InPhase>
Or to whatever fork of that you're using.
<linext>
i checked it out, they use a preview image from openscad to show the model
<linext>
they're also asking people to subscribe for $
<linext>
the free account is locked to being able to save 3 models
<linext>
i fixed the bug in 3dcustomizer.net where you'd have to manually click "render" twice, and the case where rendering would keep occurring in a loop
<InPhase>
Ah, excellent. I had those bugs happen to me a few times. :)
<JordanBrown[m]>
Without the Windows version of Python installed, cmake found the MSYS2 version. After that, I needed to set $PYTHONHOME, which for some reason wanted to be /c/msys64/mingw64:
<guso78>
can somebody setup a template python test for me ?
<guso78>
python code can be as simple as: cube().output()
guerd87 has quit [Read error: Connection reset by peer]
JoelJoel has quit [Quit: ZNC 1.8.2+deb2build5 - https://znc.in]
Joel has joined #openscad
hrberg has quit [Remote host closed the connection]
leptonix has quit [Ping timeout: 256 seconds]
hrberg has joined #openscad
leptonix has joined #openscad
<guso78>
teepee, i have added python support for the commandline in openscad.cc. Tried to minimize the impact basically copying most from MainWindow.cc
<guso78>
you can find it along with the other changes in my python branch.
guerd87 has joined #openscad
ur5us has joined #openscad
Scopeuk has quit [Quit: Scopeuk]
Scopeuk has joined #openscad
<guso78>
I found that "rotate_extrude" in openscad has an 2d offset parameter. is it the center of rotation when using the "twist" parameter ?
retrosenator has quit [Read error: Connection reset by peer]
epony has quit [Read error: Connection reset by peer]
epony has joined #openscad
qeed_ has quit [Quit: qeed_]
qeed has joined #openscad
qeed has quit [Quit: qeed]
qeed has joined #openscad
guso78 has quit [Quit: Client closed]
guerdy has joined #openscad
guerd871 has quit [Read error: Connection reset by peer]
Guest4358 is now known as Sauvin
<JordanBrown[m]>
guso78: I would drop the "file", "layer", "origin" and "scale" arguments to rotate_extrude entirely. They are legacy from before there was the general-purpose "import" mechanism, and are now better done by applying transformations to the 2D object before extruding it.
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
<JordanBrown[m]>
Similarly for the "file", "layer", and "origin" arguments to linear_extrude. (But keep "scale"; that scales the extrusion from bottom to top.)
guso78 has joined #openscad
retrosenator has joined #openscad
<retrosenator>
JordanBrown[m]: its just a threaded thing but with extremely fine detail
<retrosenator>
so I exceeded 2 million triangles or something
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
<JordanBrown[m]>
Yeah, that could get you into trouble. Assuming that you're working with conventional plastic 3D printing, remember to keep the limitations of the printer and the material in mind. There's no point to having details smaller than 0.1mm if that's the resolution of your printer - and really the resolution of the whole process is probably not even that good.
<retrosenator>
it claims 0.2mm
<JordanBrown[m]>
the printer?
<retrosenator>
yes
<retrosenator>
but I think i need resolution better than that to achieve it
<retrosenator>
in my mind that means I want detail in the file down to 0.05mm or so
<JordanBrown[m]>
I don't think so, but I've never really thought about the math involved. I'd suggest trying significantly lower resolution, printing it, and seeing how it comes out.
<retrosenator>
it didnt come out great
<JordanBrown[m]>
But how much of the not-great-ness is due to the resolution of the model, and how much to the resolution of the physical processes involved?
<retrosenator>
its a good question
<JordanBrown[m]>
You might try modelling a small piece of your overall model at different levels of detail, to see whether the end result changes.
<retrosenator>
yeah
<JordanBrown[m]>
For instance, I would expect the plastic to be squashed into an oval, with the Z dimension being the layer height and the X/Y dimension being the extrusion width. For a layer height of 0.1mm, that seems like it means that the side of a model will have 0.05mm of difference from the vertical center of a layer to the midpoint between two layers.
<JordanBrown[m]>
For a layer height of 0.1mm and an extrusion width of 0.4mm, I'd expect the cross section to look something like this:
<JordanBrown[m]>
and that's if everything in the picture is perfect.
<JordanBrown[m]>
You might try asking on the mailing list what people have found to be the minimum practical thread dimensions. My guess is that, by screw standards, the answer is very coarse.
<Scopeuk>
I'm sure someone was playing around and made it down to m2 or m3, not sure if that was with a 0.4 or 0.2 mm nozzle
<JordanBrown[m]>
Way coarser than any metal machine screw, and probably way coarser than manufactured plastic machine screws.
<JordanBrown[m]>
OK, when I'm thinking of metal machine screws I'm thinking of small ones; I wasn't thinking of big ones and they have big pitch.
<JordanBrown[m]>
So m2 and m3, coarse, are a pitch of 0.4 or 0.5mm.
<JordanBrown[m]>
I don't work in that kind of fine detail, but... that seems like it might be possible but would be really hard.
<JordanBrown[m]>
Anyhow, again, I'd ask on the mailing list. There are some pretty high-end designers there who can tell you what's practical and how to achieve it.
guso78 has quit [Quit: Client closed]
guso78 has joined #openscad
guso78 has quit [Client Quit]
J23591 has quit [Quit: Client closed]
J23591 has joined #openscad
<retrosenator>
Scopeuk: i'm trying to use SLA not FDM
<retrosenator>
anyway openscad finishes eventually
<retrosenator>
the problem is it needs to use swap memory
<InPhase>
"Why do you need 64GB of RAM?" people ask. OpenSCAD... OpenSCAD.
<InPhase>
Scopeuk: I printed M3 with 0.4 a few times in tests, and it's barely representing the proper shape. Trying to couple printed M3 inner and outer threads is approximately a non-starter with 0.4 nozzles.
<InPhase>
But you can just barely couple them with metal on the other end, but only because the metal will "correct" the defects.
<Scopeuk>
Fair enough, yeh tap/thread with hardware
<InPhase>
M4 is the minimum that's really viable with that printer setup profile.
<InPhase>
And M5 comes out really good (if things are all calibrated properly).
<InPhase>
Better resolution should approximately scale accordingly as long as things are comparably stable and there are no added issues with plastic shrinkage type effects.
ur5us has joined #openscad
<InPhase>
For thermal deposition, the plastic tendency to self-adhere will tend to preferentially close up holes more than should be done as small holes have a high curvature.
<InPhase>
The same can happen in between layers on the inner portions of outer threads.
J23591 has quit [Quit: Client closed]
J23591 has joined #openscad
ur5us has quit [Ping timeout: 260 seconds]
Teslamax_ has joined #openscad
Teslamax has quit [Ping timeout: 246 seconds]
Teslamax_ has quit [Client Quit]
<J23591>
Scopeuk for such small threads you can use lower layer heights .. or print the thread horizontal (cutting the thread till the core diameter) i made working M1.5 threads that way
<peeps[laptop]>
lf94, sounds like you want a knob more than a nut. you can embed standard hardware nut in a knob such as my design https://www.thingiverse.com/thing:1562501
<othx>
peeps[laptop] linked to "Smooth n' Curvy Star Knob (Customizable) by thehans" on thingiverse => 4 IRC mentions
<JordanBrown[m]>
And for more BOSL2 stuff see the right sidebar under “threaded parts”
<lf94>
I do need a knob but a very very tiny one
<lf94>
and I dont have the necessary hardware (nut) to embed it in
<lf94>
it's literally just to stop my leveling knobs from jostling aronud.
<peeps[laptop]>
i have used that design for putting knobs on m3 size nuts
<peeps[laptop]>
or, rather, I guess it was m3 hex bolt heads
<peeps[laptop]>
same deal either way
<JordanBrown[m]>
Embedding a bolt is tricky. What did you do, support it so that the shaft could point down while being free?
<JordanBrown[m]>
Or do it sideways?
<peeps[laptop]>
hmm? there is a hex recess in the top of the knob, and a hole for the shaft. bolt head press-fits into hexagonal hole from the top of the knob
<JordanBrown[m]>
Ah, so only press fit, not captured.
<JordanBrown[m]>
Got it.
<J23591>
"capture by 3D doodler pen"
milza has joined #openscad
<peeps[laptop]>
i guess if you wanted to go nuts (heh), you could put a slot for a perpendicular nut to slide in, and a grub screw through that to tighten onto the side of the main bolt-head/nut
ur5us has joined #openscad
<peeps[laptop]>
but you'd need a bit more meat on the part compare to press fitting
Teslamax has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
<lf94>
include </home/lee/others/BOSL2/screws.scad>
<InPhase>
This is so grepping my log is easier next time. Hi future me! Manifold is a future possible alternative to CGAL. https://github.com/elalish/manifold