<lf94>
for day 2, is each cube representing an angle?
<lf94>
I dont think so
<lf94>
So it's condensing at the top
<InPhase>
Ah, the Armadillo.
<InPhase>
I didn't check on the full order of these.
<InPhase>
lf94: It's a y-loop rather than an angle loop.
<lf94>
What do you mean y-loop
<InPhase>
r = 20; ... for (y = [-r : 0.3 : r])
<InPhase>
This traces out each curve of boxes.
<InPhase>
So yes, they're spaced farther at the bottom and scrunched at the top.
<InPhase>
The cubes are 0.6, so twice as big as the step size.
<InPhase>
The source is in the magnification icon if that wasn't clear. We tried to make that an intuitive spot for it. :)
<lf94>
Curv is disappointingly failing hard at this many unions.
<lf94>
Yeah, I'm also trying not to look at the source
<InPhase>
Ok.
<InPhase>
This one is a particularly fast preview for OpenSCAD. Although the render takes a couple minutes it seems.
<InPhase>
The preview is about 1/8th of a second.
<lf94>
Yeah, this is not good.
<lf94>
This is actually very bothersome to me
<InPhase>
I'll confess I expected the result, because these are systems built around two styles of trade-offs, and OpenSCAD has the advantage of these being designs that seemed like a good idea under the OpenSCAD system. :)
<InPhase>
But it's still a worthy exercise to see which go which way and for what reason.
<InPhase>
Today's is 9 lines of code of the most basic approach of tossing some built-ins around in a way that somebody thought was pretty, probably while they were playing around with learning the language.
<lf94>
No no, this is an excellent stress test.
<lf94>
I'm frustrated, but this is good frustration
<InPhase>
9 lines of pretty code in any sdf approach under the same scenario would probably be some curvy thing.
<lf94>
I think it's Curv's code generation.
<lf94>
Not the actual shader produced
<lf94>
Going to inline everything now
<InPhase>
lf94: I wonder if you can speed things up by artificially imposing bounding boxes?
<InPhase>
lf94: This will be a disaster to compute if it has to do some min/max thing for all cube surfaces for the whole render, but perhaps you can segment with conditionals to divide out each arch.
<InPhase>
I'm not familiar with the language yet, and waiting for your designs to see how it's done, but if that's possible I bet it's 40x faster.
<lf94>
bounding boxes are included in every shape; yes, it's min/max all the way down.
<InPhase>
But the min/max is the problem. Do you get what I mean about segmenting?
<lf94>
The point is to recreate the exact same shape though. I can get it to look like it, but not exactly it. That's a problem.
<lf94>
Yes
<InPhase>
Ok.
<lf94>
Talking through it with Doug (Curv author)
<lf94>
Hoping he can give me some hope here.
<InPhase>
teepee: Presumably you're asleep at 4am, but did those css edits you made that fixed the zoom box centering with different browser window aspect ratios not make it to the live site? It seems to be back to the older behavior.
<lf94>
fyi the site is not usable on mobile. you probably already know
Jack2178 has joined #openscad
arebil has joined #openscad
<InPhase>
Works fine for me on my android.
<InPhase>
Are you iphoning it?
Jack21 has quit [Ping timeout: 256 seconds]
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
snaked has quit [Ping timeout: 252 seconds]
snaked has joined #openscad
TheCoffeMaker has quit [Ping timeout: 265 seconds]
TheCoffeMaker has joined #openscad
arebil has joined #openscad
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
PovilasCNC has joined #openscad
arebil has joined #openscad
GNUmoon has quit [Ping timeout: 276 seconds]
GNUmoon has joined #openscad
ferdna has quit [Quit: Leaving]
lastrodamo has joined #openscad
<knielsen>
cbmuser: The status of the new member process is listed as "Approved", so it seems to have been completed by now, which sounds great
<teepee>
nice, congratulations
ochafik has joined #openscad
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
<gbruno>
[github] bakerjvb opened issue #3990 (Rendering individual Polyhedon OK but multiples gives "mesh not closed" error: solved with hull() ). https://github.com/openscad/openscad/issues/3990
arebil has joined #openscad
<gbruno>
[github] t-paul edited issue #3990 (Rendering individual Polyhedon OK but multiples gives "mesh not closed" error: solved with hull() ). https://github.com/openscad/openscad/issues/3990
<gbruno>
[github] t-paul closed issue #3990 (Rendering individual Polyhedon OK but multiples gives "mesh not closed" error: solved with hull() ). https://github.com/openscad/openscad/issues/3990
<Scopeuk>
winding order bits again, I do wonder if maybe thrown together could add the point and surface numbers onto the render (obviously there would be limitations)
<Scopeuk>
seeing that surface 3 is white and has points 1,5,8,3 on its edges might make debugging it easier (granted this could probably be done in scad code, at least for the points
<Jack2178>
there was this right click thingy to see what geometry caused the pixel (which is not working) but that could be used to display the face or ?
<Jack2178>
/nick Jack21
<Scopeuk>
not sure what I would expect it to do in that case
Jack2178 is now known as Jack21
<Jack21>
show the faces or points of that face
<Jack21>
maybe if it spot a wrong orientation - why not correct it directly
othx has quit [Ping timeout: 256 seconds]
othx has joined #openscad
<Scopeuk>
Jack21 generally ignoring the user leads to problems
<Scopeuk>
it could be that we flip a face but actually one of the points is wrong
<Scopeuk>
producing a nicely worded warning would be a nice step though
<Jack21>
oh yeah definately with a warning - just that "not closed" is missleading
<Jack21>
i meant in this case - if faces are missing this is for sure correct
<Jack21>
maybe something like face [2] has wrong orientation
<teepee>
I'm not sure we have that much detail information
<teepee>
that would be nice, but may need quite some extensive extra logic
<dTal>
lf94: if you're playing with SDFs be sure to check out libfive as well
<teepee>
IIRC it's already using that :)
<dTal>
What is? Not curv
<teepee>
yes, but lf94 mentioned something about libfive being used for meshing, if I'm not mixing up things completely
Guest23 has joined #openscad
Guest23 has left #openscad [#openscad]
Guest23 has joined #openscad
<Guest23>
tgghttjàpdhg
<Guest23>
CHE SAMANTA NECESITO COMPRAR UNA CASA
<Guest23>
hi samanta
Guest23 has quit [Client Quit]
<teepee>
good talk :)
<lf94>
teepee's got it :p
ochafik has quit [Remote host closed the connection]
<Jack21>
Ü
<Jack21>
translates to : CHE SAMANTA I NEED TO BUY A HOUSE
ochafik has joined #openscad
ochafik has quit [Ping timeout: 252 seconds]
<lf94>
lol.
arebil has quit [Ping timeout: 245 seconds]
arebil has joined #openscad
ochafik has joined #openscad
<lf94>
InPhase: so the unions are only slow because they are *always* inlined. This is fixable though
<lf94>
As an alternative until it's fixed by Doug, I'm creating a "pixelated" sphere, which mimics the same aesthetic :)
<lf94>
In fact I should be able to copy the exact same stepping that happens.
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
<InPhase>
lf94: So Doug has in mind a generalized optimization to curv from this?
mhroncok has quit [Quit: Leaving.]
mhroncok has joined #openscad
<lf94>
Yep :)
<lf94>
Well I think he's had it in mind for awhile.
<lf94>
For sure though my concern will bring it to the forefront of his mind
<lf94>
I think this is a huge issue
ochafik has quit [Remote host closed the connection]
Jack21 has quit [Quit: Client closed]
Jack21 has joined #openscad
ochafik has joined #openscad
arebil has joined #openscad
<JakeSays>
what's new in the land of shapes and sizes?
<InPhase>
lf94: Sounds good. :) This is the sort of ""competitive"" mutual strengthening open source does well.
<Scopeuk>
https://www.seeedstudio.com/Seeed-XIAO-BLE-Sense-nRF52840-p-5253.htmlis interesting, pricey for a micro but its a tiny board with acelerometer/gyro (6 dof) blue and a lipo charge/management chip. with the right firmware it could probably do a passable impression of a 6dof mouse
mhroncok has quit [Quit: Leaving.]
peeps[zen] has joined #openscad
<teepee>
right and with the microphone you can do click by voice :)
<Scopeuk>
"NO THE OTHER LEFT!"
peepsalot has quit [Ping timeout: 252 seconds]
<teepee>
exactly!
peepsalot has joined #openscad
peeps[zen] has quit [Ping timeout: 252 seconds]
<Scopeuk>
I'm not sure how I feel about my peripherals knowing I'm yelling at them
<teepee>
just upload a french language pack so they can't understand you ;-)
<InPhase>
Allegedly English profanity is a near universally understood sub-language.
<InPhase>
So make sure it's a G-rated language pack.
<Scopeuk>
I'm sure it helps that its universally harsh sounds and single syllables
<Scopeuk>
or alternatively, detect harsh language, reduce sensitivity :P
<othx>
ali1234 linked to YouTube video "Ameca Humanoid Robot AI Platform" => 1 IRC mentions
<ali1234>
something about how the neck deforms just doesn't seem right to me
<lf94>
yeah that's what I focused on too.
<ali1234>
lol the other videos on their youtube channel
<ali1234>
yeah i'm 99% sure this is satire/art
<ali1234>
"real human hair"
<Jack21>
from the website you can order that thing to be at your venue
ochafik has quit [Remote host closed the connection]
<lf94>
ali1234: I'm pretty sure it's real
<lf94>
real human hair is not a big deal
<lf94>
People donate their hair all the time to make wigs
PovilasCNC has quit [Read error: Connection reset by peer]
ochafik has joined #openscad
<JakeSays>
Scopeuk: i did a pre-order from seeed for an xiao rp2040 back in july. still waiting.
mhroncok has quit [Quit: Leaving.]
ochafik has quit [Remote host closed the connection]
<Scopeuk>
JakeSays fair enough. I tend not to pre order, I have some original XIAO's and they worked great for the couple of bits I played with
<Scopeuk>
I have a day dream remote that I half patched in to let you use it to manipulate the object in the open scad window
<Scopeuk>
I got rotation working at about 85% but never got translation working well. that's a more controllable endpoint and class compliance would make things easier
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
GNUmoon has quit [Ping timeout: 276 seconds]
Guest94 has joined #openscad
<Guest94>
cool this open SCad
<Guest94>
May I ask which language and tools it is written?
<teepee>
OpenSCAD itself is written in C++
<Guest94>
Thx mate
<Guest94>
Was looking for a design program for 3-D printing. That it is!!!
<Guest94>
Whar= IDE is it
<teepee>
yep, that's certainly a topic where it works quite well
<Guest94>
Looks like single source for several OS?
<teepee>
yes
<Guest94>
Is that QT or is that a secret :-)
<Alicia>
it is Qt
<teepee>
no secrets (well, except the doors of the advent calendar) - yes, Qt and the editor is QScintilla
<Guest94>
ah ok, they have a community edition
<Guest94>
The real thing got so expensive no mere mortal can afford that anymore
<Guest94>
I like especially that it does NOT have an installation procedure
<teepee>
luckily there's the open source edition, hopefully that continues to be supported
<Guest94>
Have a nice evening and see you again for sure
<teepee>
thanks, you too
Guest94 has quit [Quit: Client closed]
Jack21 has quit [Quit: Client closed]
Jack21 has joined #openscad
<JakeSays>
Scopeuk: yeah i usually don't pre-order either, but i figured what the heck. the xiao's are pretty slick. i've got 7 of them running in various devices