<linext_>
sliding in a piece of lasercut acrylic into the slot
ToAruShiroiNeko has quit [Remote host closed the connection]
Andy-pants has joined #openscad
ToAruShiroiNeko has joined #openscad
peeps[zen] has joined #openscad
To_Aru_Shiroi_Ne has joined #openscad
peeps has quit [Ping timeout: 255 seconds]
ToAruShiroiNeko has quit [Ping timeout: 255 seconds]
epony has quit [Remote host closed the connection]
<linext_>
is it possible to set the max length of a string?
To_Aru_Shiroi_Ne has quit [Ping timeout: 255 seconds]
ToAruShiroiNeko has joined #openscad
ToAruShiroiNeko has quit [Excess Flood]
Andy-pants has quit [Ping timeout: 255 seconds]
Andy-pants has joined #openscad
<gbruno>
[github] JonMcLin opened pull request #4537 (Implement ExportPdfOptions Dialog with paper selection and more (#4143, partial #649)) https://github.com/openscad/openscad/pull/4537
fling has quit [Remote host closed the connection]
fling has joined #openscad
Andy-pants has quit [Ping timeout: 255 seconds]
Andy-pants has joined #openscad
Andy-pants has quit [Ping timeout: 248 seconds]
LordOfBikes has quit [Ping timeout: 248 seconds]
LordOfBikes has joined #openscad
<Av8r>
Any thoughts on the the Linux (Ubuntu) IDE most friendly to OpenSCAD dev? (C++/Qt/BOOST) I'm ready to be done with using Gedit...
<InPhase>
vim
<InPhase>
If you really want to pack it into an IDE though, Qt Creator works pretty well for cmake-based Qt programs.
<InPhase>
But you can easily do command line builds with cmake. It's extremely simple.
<Av8r>
I was thinking more along the lines of integration with GDB. I used Qt Designer for the UI, but the IDE felt rather heavy.
<InPhase>
Qt Creator does do an okay job of debugger use if you want such a thing.
<InPhase>
For my standard Linux build process, I use this for clean builds: https://bpa.st/ZQHB6 and this for incremental builds: https://bpa.st/3JU3S By using shell scripts, I lock in my standard compile settings, and then can use them across many repository clones on different PRs.
<Av8r>
Probably Git integration would be a plus, but I just don't know what that would look like. My CM integration experience with IDEs was Visual Studio 25 years ago on Windows NT, and then CVS.
<InPhase>
I use gitk for a git history browsing gui, and then add/commit/push/pull/status/diff and such I do by command line.
* teepee
uses NetBeans
<Av8r>
I had no idea tcl/tk was still being maintained. Wow.
<teepee>
tcl is still used heavily in semiconductor production environment
hyperair has quit [Ping timeout: 255 seconds]
<teepee>
C++ support in NetBeans is at a crossroad though, I hope C++ will keep going
<teepee>
there's a pretty good plugin from ancient version 8 that is still supported in latest netbeans but it's stuck at about c++14/c++17
<Av8r>
teepee: was wondering about that. Got the sense it was focused on internet stuff, and from the name "Java" (like Eclipse).
<teepee>
unfortunately only parts of that were transfered from oracle to apache
<teepee>
so supposedly there's a new full version coming, but at this point it's just called C++ Light with limited features
<Av8r>
Hmm. Both tcl/tk and netbeans trace through Oracle. Bad omen...
<teepee>
yes, but they did not totally mess up the hand-over to Apache as with other projects
<teepee>
the old plugin is still good with full gdb integration and all the normal stuff
<Av8r>
In any event, I just need to get my feet wet in a modern IDE and C++. I liked the spyder IDE with Python, although anaconda seems to have made it unusable on Ubuntu. I like SCiTE as an editor. Thought Qt Creator may be too Qt-centric (given the need to segregate Qt code from core code). Code::Blocks is wx based. That could conceivably
<teepee>
plus it's impressive that the old plugins still just work, they traditionally have an extremely good sense of API handling
<Av8r>
be problematic with Qt (from a bias perspective). Kdevelop may have concerns similar to with Qt Creator. So really: just looking for "try this, it works well with the code structure of OpenSCAD".
<teepee>
I've never used CLion, the JetBrains IDEs are usually quite good
<teepee>
at work I have a PyCharm license, maybe I can check if that license covers CLion too
<Av8r>
That's good. Don't like brittle code! And being a cranky old man, I'm looking for something which isn't likely to require me to redesign my workflow on a regular basis... I want to dedicate my learning skills to whatever problem I'm trying to solve rather than the tools I'm using to get there!
fling has quit [Ping timeout: 255 seconds]
<Av8r>
I've downscoped my sense of adventure...
fling has joined #openscad
<Av8r>
Perhaps the most important part of this discussion: No one has said "don't use <X>", which suggest there are none which clearly should be avoided.
<teepee>
I try to avoid Eclipse, but I would to force that on others :)
<teepee>
mostly due to bad experience in the past running it on Linux - might be different when using it on Windows
<Av8r>
I shoujld probably start with Qt creator, since it's installed. I guess the main concern would be with overlaying it on top of my existing Openscad directory/project. I'd need to figure out how to make it happy with the existing project.
<Av8r>
There is probably a Qt or Stackexchange page on that; I didn't search for something like that yet...
<teepee>
should be fine, I think. I've only used it for a couple of hours but I assume it should handle the existing project ok
<Av8r>
I'm always paranoid about losing work. I had a great CVS workflow decades ago; I'll need to recreate an equivalent.
<Av8r>
(Maybe not "great", but I felt confident I wouldn't lose my history/work.)
<teepee>
regular local git checkins could be able to handle that topic quite well
<teepee>
it's always possible to polish that before publishing that to the server
<Av8r>
I seem to recall similar Eclipse challenges on Linux, but I don't think I've run it on Linux since before it was called "Eclipse". As I recall (vaguely) IBM developed it as a "Visual Java" environment, or somehting along those lines.
<Av8r>
It was impressive, but unusable. That was a LONG time ago.
<Av8r>
Can't remember the name. That was my first exposure to Java. Maybe they called it "VisualJ"?
<teepee>
Something like that, never used it. I started with "Forte" at some point which was the predecessor of NetBeans
califax has quit [Ping timeout: 255 seconds]
Andy-pants has joined #openscad
<teepee>
I don't remember what I used for Java 1.1, maybe just text editing with Emacs :)
<Av8r>
It was "VisualAge" - found a SUN article on it...
<Av8r>
Eclipse grew out of that as I recall.
<Av8r>
That was 1997.
<Av8r>
Apparently they are still marketing a "VisualAge"?
califax has joined #openscad
<teepee>
oh!
<teepee>
"Most of the members of the VisualAge family were written in Smalltalk no matter which language they supported for development."
<Av8r>
I was never able to Grok EMACS. I've used it, but it felt heavy.
<Av8r>
I was quite comfortable with VI. Now there's this newfangled "vim". Which seems the same, except with color-coding!
<Av8r>
(that means I could figure out how to use it without checking the man pages or buying an O'Reilly book...)
<teepee>
yeah, the number of Java IDEs at the time was crazy, and some crazy expensive too
J232690 has joined #openscad
<Av8r>
Need to sign out for the night, thanks!
J2326 has quit [Ping timeout: 260 seconds]
teepee_ has joined #openscad
teepee has quit [Ping timeout: 255 seconds]
teepee_ is now known as teepee
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
Andy-pants has quit [Ping timeout: 255 seconds]
<InPhase>
Av8r: Syntax highlighting and indentation support with vim are what really make it suitable as a code editor. But vim is no longer the newfangled variant. Now there is neovim.
<InPhase>
Av8r: If I were starting fresh, I might start with neovim. But for now my brain is pretty locked into standard vim after 25 some years of using it as a primary code editor.
<InPhase>
Av8r: Also! You can use vim as an external editor for scad code using the support for that within OpenSCAD. And there is a syntax highlighting support file you can grab for vim that adds OpenSCAD syntax highlighting.
fling has quit [Remote host closed the connection]
fling has joined #openscad
ToAruShiroiNeko has joined #openscad
epony has joined #openscad
J232690 is now known as J23
<J23>
linext_ you can check the length and use assert or use a function to truncate after n
Vasiliy475 has joined #openscad
fling_ has joined #openscad
fling has quit [Ping timeout: 255 seconds]
TheAssassin has quit [Ping timeout: 255 seconds]
GNUmoon2 has quit [Ping timeout: 255 seconds]
TheAssassin has joined #openscad
fling_ is now known as fling
vasiliy has joined #openscad
vasiliy has quit [Client Quit]
GNUmoon2 has joined #openscad
VasiliyB has joined #openscad
<J23>
linext_ e.g. function String(in,length=0,out="")= length>len(out)?String(in,min(len(in),length),str(out,in[len(out)])):out;
VasiliyB has quit [Client Quit]
VasiliyB has joined #openscad
Andy-pants has joined #openscad
Andy-pants has quit [Ping timeout: 248 seconds]
hyperair has joined #openscad
califax has quit [Remote host closed the connection]
<J23>
getting points from an existing object is not possible (there is something in development for rendering an object into points)
guso78 has joined #openscad
<guso78>
i just changed the outrput format of my facetracer to obj to circumvent the issue, i dont yet have tesselation.
<guso78>
now i want to use the obj file with openscad and i realize, that it cannot handle.
<guso78>
why ? is it that its challenging to tesselate a polygon with 3d points rather than 2d points ?
Vasiliy475 has quit [Ping timeout: 260 seconds]
<teepee>
openscad can automatically tesselate when needed
foul_owl has quit [Ping timeout: 252 seconds]
<teepee>
e.g. polyhedon allows to give faces with more than 3 vertices
<guso78>
yes, but tesselation is needed with my python script right now ... . as tesselation is not the main focus for my facetracer, i try not to do it.
<guso78>
need an graphical exporrt to debug the script
<guso78>
now changing output to OFF format.
<guso78>
BTW are there any plans to support OBJ format with openscad. it appears to be an easy format
<guso78>
next challenge are cuts, where the cutface cuts edges of an object
<guso78>
another challenge will denitely be coplanar faces
<teepee>
how does this compare with what ochafik is doing?
<guso78>
ochafik integrates an existing lib with openscad. manifold is developped by emmett.
<teepee>
oh, I thought you use manifold too, so that's new code based on that paper?
<guso78>
until i get good proof that manifold can actually fix intersecting polygons, i will try to do it by myself.
<guso78>
no its new python code where i implement my understanding of a line tracer for 3d world
<guso78>
(i know line tracer algorithm from assura, vampire, pvs, calibe, dracula ....)
<guso78>
and it do it in python yet, because no compilation is needed
<teepee>
cool, mesh fixing is still a big topic. on the mailing list there's someone who wrote a mech fixer but it's under a license that would not work with openscad unfortunately
<teepee>
I guess admesh would be usable as library but using it as external tool did not give confidence anywhere near 100%, it does fix some issues, but not all classes that cause problems
<guso78>
haha, teeepee, you got 3 guesses why i am so eager to achieve a working "mesh fixing"
teepee has quit [Quit: bye...]
<guso78>
of course i will adhere to the paper to benefit from already solved problems
<JordanBrown[m]>
I have a mesh fixing module. The only problem is that the fix is very low quality. You just say fix() import(whatever) or fix() polyhedron(whatever).
<JordanBrown[m]>
Here it is: module fix() cube();
<JordanBrown[m]>
Like I said, the fix is very low quality.
<JordanBrown[m]>
But it always works.
<JordanBrown[m]>
And it's very very fast.
ur5us has quit [Ping timeout: 260 seconds]
teepee has joined #openscad
<guso78>
JordanBrown[m]. whats the diffeence between low quality "Working" and high quality "working" ?
<JordanBrown[m]>
How high the quality is, of course.
<guso78>
is it an openscad module ?
<JordanBrown[m]>
I gave the source above.
<JordanBrown[m]>
module fix() cube();
<JordanBrown[m]>
For any input, it gives an geometrically valid approximation of the input.
<JordanBrown[m]>
Unfortunately, for most inputs the approximation is not very similar to the intended shape.
<guso78>
just curious. is it possible to share ?
<JordanBrown[m]>
It's a joke.
<guso78>
aha,
<JordanBrown[m]>
The entire module is "module fix() cube();".
<JordanBrown[m]>
For any input, it yields a 1x1x1 cube.
<JordanBrown[m]>
It's a really poor approximation for most inputs.
<guso78>
haha if its not what the user expects by fixing geometry, neither InPhase , would probably be happy with the result
<guso78>
and i need to have very good post processing for my offset haha
<JordanBrown[m]>
Perhaps not. Some people are picky that way.
<JordanBrown[m]>
But it can do something that almost no other fixer can do: it can even fix nonexistent files.
ur5us has joined #openscad
<guso78>
haha, this could even be suspecious to be true
<guso78>
and it can even work on non-causal input
<guso78>
i;ll have a look to admesh now probably
<guso78>
admesh does all the things with an stl file which can be done with "simple calculations" but unfortunately according to its featrues, none of them involves CSG.
<guso78>
this is definitley needed for my mission
Andy-pants has joined #openscad
epony has quit [K-Lined]
<teepee>
heh, someone got k-lined again
<teepee>
yes, admesh is "just" for a single mesh
<JordanBrown[m]>
What does "single" mean here?
<teepee>
meaning the canonical output if the input is fine is the same as the input, no operations performed CSG or otherwise