kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
kyki has quit [Remote host closed the connection]
kyki has joined #openscad
kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
kyki has quit [Remote host closed the connection]
kyki has joined #openscad
kyki has quit [Remote host closed the connection]
kyki has joined #openscad
myosotis has quit [Quit: myosotis]
kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
kyki has quit [Remote host closed the connection]
kyki has joined #openscad
kyki has quit [Ping timeout: 272 seconds]
kyki has joined #openscad
J221168 has joined #openscad
J2211 has quit [Ping timeout: 256 seconds]
kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
k3gk has joined #openscad
kyki has quit [Ping timeout: 240 seconds]
k3gk has quit [Ping timeout: 272 seconds]
Junxter has quit [Ping timeout: 240 seconds]
kyki has joined #openscad
kyki has quit [Ping timeout: 240 seconds]
kyki has joined #openscad
kyki has quit [Remote host closed the connection]
kyki has joined #openscad
kyki has quit [Read error: Connection reset by peer]
kyki has joined #openscad
k3gk has joined #openscad
kyki has quit [Remote host closed the connection]
ferdna has quit [Quit: Leaving]
k3gk has quit [Quit: Leaving]
kyki has joined #openscad
kyki has quit [Client Quit]
arebil has joined #openscad
Junxter has joined #openscad
lastrodamo has joined #openscad
peepsalot has joined #openscad
use-value has quit [Quit: use-value]
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
arebil has joined #openscad
kyki has joined #openscad
kyki has quit [Quit: Leaving]
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
Junxter has quit [Ping timeout: 240 seconds]
pie_ has quit [Ping timeout: 256 seconds]
pie_ has joined #openscad
pie_ has quit [Ping timeout: 256 seconds]
pie_ has joined #openscad
Wolf480pl has joined #openscad
linext_ has quit [Read error: Connection reset by peer]
linext has joined #openscad
<OlivierChafik[m]>
teepee: I've made a breakthrough with the WASM stuff... just rolled out my own filtered numeric type with larger intervals and it seems to work
<teepee>
OlivierChafik[m]: nice, do we still need that secret "round to nearest" flag, I think that's not upstream either
<teepee>
for font support there's something in the example folder, but I did not try that yet
<OlivierChafik[m]>
oh I'm not sure, where is the rounding flag right now?
<OlivierChafik[m]>
(you mean the cgaljs patch?)
<OlivierChafik[m]>
I've just earned a breakfast break ✌️
<teepee>
no, cgaljs was patching that, but I found that secret flag in one of the issues we discussed some days ago
<OlivierChafik[m]>
(Also the viewer has rendering glitches that I suspect come from its STL import logic; hope it's not the stock Three.js import, I'll probably try and follow this tutorial to get our own renderer https://tonybox.net/posts/simple-stl-viewer/ )
<teepee>
on it's own it did not help
<teepee>
but maybe it's still sensible to have it
<OlivierChafik[m]>
ah ok, well, maybe it could become useful if we still see some asserts
<teepee>
I have to check my test builds if there's anything open for a PR
<OlivierChafik[m]>
please try and break the new version ;-)
<teepee>
I might, just need to take it a bit slower today, yesterday I was out with a bit of a headache so trying to not bring it back today :)
<teepee>
oh, so the +0/-0 was what caused the macos tests to fail?
<OlivierChafik[m]>
(once/if it gets in, I'll create a cgal subfolder to clean stuff up a bit as discussed the other day)
<teepee>
very cool it's working now
<OlivierChafik[m]>
yes, remeshing resulted in some -0 normals (and/or vertices, can't remember), sometimes haha
<OlivierChafik[m]>
I'm sure it can be sped up quite a bit, e.g. I'll look into using Surface_mesh builtin property maps (vectors instead of unordered_map), etc, but best as a follow up
<teepee>
yeah, making PRs too big has proven to be quite problematic, there's still a number of things hanging out there for years that would be nice to have
<teepee>
so splitting stuff up like you do is certainly the better way to go
<OlivierChafik[m]>
although the remeshing is already a beast to review, sorry :-D
<OlivierChafik[m]>
alright, off i go, just wanted to share the update, ttyl / have a great day!
<teepee>
thanks, you too
J221168 is now known as J22
<J22>
did the auto intend changed .. it is not working as i am used to
<J22>
indent .. should increase after a curled bracket or?
<J22>
and the closing bracket should be unindeted
<teepee>
hmm, not really
<teepee>
maybe it's a side effect of the new lexer, that could have caused some different behavior I think
<J22>
also if you have empty lines they always get indented to the last indent even if there are 0 indent lines before
<teepee>
right, that's annoying. maybe there's some flags in the lexer to change that behavior, otherwise I have no idea why it would have changed
<J22>
just checked with an older version and that does the auto indent after brackets
<J22>
and removes one for the closing curled bracket
<teepee>
hmm, not seeing the code for that yet. the lexer has some folding stuff, but not so much about indent
<J22>
interesting if you now disable auto indent .. everything starts at the beginning of the line
<teepee>
well, it probably is there somehow :)
<J22>
hmm ok auto indent is also that new lines start at the same indent ( i thought auto indent was that bracket behavior only)
<J22>
but increasing the indent after a curled bracket was useful - at least for me
<teepee>
ah, found the indentation code in qscintilla and it is calling the lexer as I assumed
<teepee>
so it will need some digging into the deeper layers to see what's different now
<teepee>
I guess the old one did get some sort of default c++ indentation behavior and now we are getting the "custom lexer" default
use-value has joined #openscad
dbovey has joined #openscad
<dbovey>
Hi, I have a "parser error: unterminated string". It must be something very stupid, but for the life of me I cannot find what is wrong with my scad file.
<dbovey>
}
<dbovey>
I could not fina any indication on any openscad forum or doc about this.
<peepsalot>
sounds like a double quote " somewhere without a matching one to close the string
<peepsalot>
syntax highlighting should also help with that I would think
<peepsalot>
dbovey: crtl-f: "
<peepsalot>
*ctrl
<dbovey>
yup, that was it, [shame]. thanks for the quick answer.
<InPhase>
Syntax highlighting in the editor helps.
<J22>
Matt11 but looks nice .. instead of cutting the slider you can just make it smaller and calculate the twist per height by dividing the inners twist / height * height_slider
<J22>
instead of using slices just use $fs=.5;
<J22>
$fa=1; ( or $fn) ..
<Matt11>
Thanks for all the advice this is very helpful!
<Matt11>
As for the twist per height part, I did think about doing that, but then I would have to apply a initial rotation that I would have to calculate, which wouldn't be too hard to do I just thought it would be easier to generate the whole thing and cut off stuff on the top and bottom. But performace wise the way you mentioned would probably be quite a
<Matt11>
bit better
<J22>
and as this is a fidget toy .. chamfer the edges
<J22>
.. do you plan to print this as one part?
<Matt11>
Yeah I've printed several of various sizes already as I have been testing this script, prints quite nice
<J22>
with supports i assume
<Matt11>
So supports between the bottom base and the outer spinning part, no other supports
<J22>
you can have much lower clearance if printed separately .. but probably you don't need that tight fit for it
<Matt11>
Ah okay that is a good idea, print it separately, then just slide it on before it prints the top cap
<Matt11>
So you mentioned " -+ offset to get rounder edges" what exactly did you mean by that? And how would I do that?
<J22>
if you make an intersection of the center part with a mirrored version (or opposit twisted) you can use two sliders running in opposite directions
<J22>
offset(+1) offset(-1) circle(20,$fn=8);
<J22>
make sure you define $fa $fs (or $fn) globally in that file
<Matt11>
Awesome! Thanks so much for your help, I got to head out for a bit but if you leave any more advice I'll be sure to check the chat logs
<peepsalot>
teepee, InPhase: I'm considering moving GUI/Qt-dependent files into a new subdir of "/src". does that sound okay? what would be a preferred name: gui, GUI, qt, Qt, something else?
<teepee>
gui sounds good to me
<peepsalot>
looks like about 85 files in src are GUI specific. that's not including src/parameter and src/input... which I think should also probably move under the same path
<InPhase>
peepsalot: Yeah, not qt or Qt because those directories might already be used by Qt.
<InPhase>
I mean, elsewhere, but by the time you add enough include paths, things get messy sometimes.
<InPhase>
gui sounds reasonable to me.
<peepsalot>
ok, lets see how this goes...
<InPhase>
*15 minutes later* * peepsalot has flipped the desk over.
<peepsalot>
hehe, i doubt it will be much hassle, but we'll see
<teepee>
the only extra hassle I would expect is updating the translation scripts :)
<teepee>
otherwise cmake should take care of things hopefully
<InPhase>
Hmm. Qt is very "leaky" into this codebase. I'm counting about 109 files plus or minus with Qt stuff in them, but not all are gui related.
<InPhase>
Like, value.h...
<InPhase>
It has a 2013 era Qt bug workaround.
Ckat has joined #openscad
<Ckat>
hello, I'm intersted in ordering some faceplates made with openscad from pcbway, would the best method be to export to dxf (its a 2d drawing), and use that? I tried viewing it with an online viewer and it seems rather insufficient as its lacking infills and stuff
<Ckat>
oh yeah, not 3d printed, lasercut/cnc rather
<InPhase>
peepsalot: This is a candidate list that appears to be doing Qt stuff. https://bpa.st/MVSA
<InPhase>
peepsalot: Although some are Qt without categorically making sense as gui, like version_check.h which just obtains some Qt versioning, and openscad.cc which is obviously top-level.
<InPhase>
peepsalot: I don't know about Network* though. That's Qt but not gui, so maybe the directory name choice locked that out.
<Ckat>
probably needed for the printer stuff
<InPhase>
teepee: Does that SparkleAutoUpdater actually get used on the mac version? That's the first I noticed the .mm :)
<peepsalot>
i ended up with 85 files, I just used this regex for qt class name convention: "Q[A-Z][a-z]"
<InPhase>
I grepped for Q and then removed a bunch of false positives manually.
<peepsalot>
some I didn't include: cache.h, version-check, openscad.cc, GeometryEvaluator, LibraryInfo.cc
<peepsalot>
Camera.h?
<InPhase>
Oh. False positive I missed removing.
<peepsalot>
GLView.h
<InPhase>
Same. I mean in my defense I did like 6 passes. ;)
<InPhase>
Ckat: Sorry for the delay. dxf export seems to be the standard approach for lasercutting approaches. But I haven't tried it myself.
<Ckat>
I mean its fine if it does work, but I dont have a dxf viewer or experience with how these files should look, I could always just try it to see what pcbway makes of it I guess
<InPhase>
Ckat: I would probably try to install a dxf viewer.
<Ckat>
as long as my general idea of just doing a flat square based drawing along x and y amounting to a usable dxf file for lasercutting/cnc
<Ckat>
I tried a couple online ones with varrying results of how much they liked the file, but sure I can try finding some other software to view it and make sure its fine
<J22>
dxf can have multiple layer for different settings - but all 2D exports from SCAD (svg, dxf ) have only one layer .. so if you need different you need multiple files
myosotis has joined #openscad
<Ckat>
I probably do if I want engraving, but I dont need that atm
<J22>
also SCAD are always closed paths ..
<Ckat>
that being an issue in what way?
<teepee>
The network stuff is for the design export to octoprint or print service, so it's pretty much GUI
<teepee>
I think Sparkle is only build for release versions
<J22>
you can't make a cut where the laser only travel once
<Ckat>
ah, not really an issue for me
<J22>
yeah it is more if you engrave pattern or text
<Ckat>
I guess I'll just fully draw it out as I was doing, export to dxf and see what pcbway thinks of it
<J22>
if it is nothing complex it should be easy to determine what is outside and inside