teepee changed the topic of #openscad to: OpenSCAD - The Programmers Solid 3D CAD Modeller | This channel is logged! | https://openscad.org/advent-calendar-2021/ | Website: http://www.openscad.org/ | FAQ: https://goo.gl/pcT7y3 | Request features / report bugs: https://goo.gl/lj0JRI | Tutorial: https://bit.ly/37P6z0B | Books: https://bit.ly/3xlLcQq | FOSDEM 2020: https://bit.ly/35xZGy6 | Logs: https://bit.ly/32MfbH5
<InPhase> teepee: I didn't get a chance to reply this morning while rushing out the door, but I had a good solid chuckle at 08 failing as an octal. I enumerated a set of explanatory theories in my head, but didn't make it to octal parsing of dates. :)
ur5us has quit [Ping timeout: 252 seconds]
<InPhase> lf94: Day 9 opened. :)
GNUmoon has joined #openscad
ferdna has joined #openscad
<Jack21> i like to print that but that is not be easy .. probably edge down
<Jack21> hmm seems i already thought about that when making the 3mf
<Jack21> it is kind of a disphenoid  https://en.wikipedia.org/wiki/Disphenoid     other names - almost regular tetrahedron  - Ü
<lf94> God damn, another mesh based one
<lf94> X)
<lf94> I could probably do it in curv without being mesh based
<lf94> What's the shape when its not a cube?
<lf94> It's hard to make out
<lf94> This will be easy with `morph`
<Jack21> not sure if it has a name (maybe i should name it)  ..  it is kind of a rounded disphenoid   so based on a tetrahedron
jonasbits has joined #openscad
oldlaptop_ is now known as oldlaptop
ur5us has joined #openscad
TheAssassin has quit [Ping timeout: 276 seconds]
TheAssassin has joined #openscad
LordOfBikes has quit [Ping timeout: 240 seconds]
LordOfBikes has joined #openscad
jonasbits has quit [Ping timeout: 252 seconds]
jonasbits has joined #openscad
ferdna has quit [Quit: Leaving]
<lf94> So it's just a ... triangular prism?
<lf94> or a triangle-based pyramid?
Jack2191 has joined #openscad
<lf94> Rendering just the points, it makes a lot more sense.
<lf94> It looks like a pillow!
<lf94> How do I animate
Jack21 has quit [Ping timeout: 256 seconds]
<lf94> The shape the points make when t=0.5 is nuts
<InPhase> Yeah. When you take out the hull() line and do animation, it's even harder to tell what's going on. It looks like one of those attempts to display a 4D shape. :)
<gbruno> [github] kintel pushed 1 modifications (Use built-in libz on macOS). https://github.com/openscad/openscad/commit/5c2a251f12157423e36e12a78579912893d3df05
ur5us has quit [Ping timeout: 252 seconds]
peepsalot2 has joined #openscad
peeps[win] has quit [Ping timeout: 256 seconds]
<JakeSays> Jack2191: there is 2170 more of you
GNUmoon has quit [Ping timeout: 276 seconds]
veverak has quit [Ping timeout: 256 seconds]
veverak has joined #openscad
GNUmoon has joined #openscad
lagash has quit [Quit: ZNC - https://znc.in]
lagash has joined #openscad
la1yv has quit [Remote host closed the connection]
la1yv has joined #openscad
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #openscad
Jack2191 is now known as Jack21
<Jack21> jakesays  need to get this multiplication urge under controle
<JakeSays> lol
KimK has joined #openscad
<Jack21> lf94  just sliced it and it is i circle that is scaled from [0,1] to [1,0] while moving along Z  (at least this is very close)
PovilasCNC has quit [Read error: Connection reset by peer]
PovilasCNC has joined #openscad
<JakeSays> my kingdom for parallel rendering
<veverak> !!!, I could ask here
<veverak> so, I have OT question for you:
<Jack21> king Jake, Sir  .. i am afraid  there is nothing you can offer
* JakeSays cries at all these idle cores
<veverak> for my robot project I made yet another mounting system for stuff (yeah, we need another generic mounting system and there is some XKCD joke viable here), I had a moment to improve it's github repository and documentation quite a bit (this is something I use for a while, so it's mostly about documenting it), mind giving me some feedback on the repo?
<Jack21> SCAD the great Equalizer where a 32-core is not faster as a dual core
<JakeSays> Jack21: unfortunately
<Jack21> i had projects where i split the model and started multiple instances to render parts (although not sure if that was faster in the end)
<JakeSays> i could dedicate 96 cores to this if it were possible
<Jack21> jakesays  wow is that on a single chip?
<JakeSays> Jack21: no. two machines 32/64 split
<JakeSays> they are about 1200 miles apart
<Jack21> veverak  looks nice, maybe a chamfer would be a nice extra  and orient the hex holes (are these for nuts?)
<Jack21> jakesays - not sure if that was meant with distributed rendering .. Ü
<JakeSays> lol
<Jack21> from what i saw the 64cores are slower on a single core compared to a quad or hex
<JakeSays> this isn't a complicated model and it's taking forever
<JakeSays> hmm. gpu rendering would be awesome
<teepee> yeah, but the most interesting part would be compute not display render
<teepee> who knows, maybe at some point we could get some student interested in working on such topics in scope of a university project
<Jack21> veverak  not sure why you need that much code for a simple part .. but there is no example and calling just `tile()`causes warnings all over the place
<othx> teepee linked to "Rubber duck debugging by JustinSDK" on thingiverse => 1 IRC mentions
<Scopeuk> with the new(ish) cache behaviour would running multiple renders across your prerequisites in independent instances and then running your final model actually pick up the cached prerequisites?
<Jack21>   `tile()`
<JakeSays> it's taking so long that my printer keeps cooling down, dammit
<JakeSays> teepee: yeah compute is what i meant
<Jack21>    `tile()`   how is that markup working?
<Jack21> jakeSays .. use peltier heater and reverse polarity
<JakeSays> huh?
<JakeSays> i don't think peltier's can get hot enough for beds
<JakeSays> damn. 38 minutes
<Jack21> 83°C
<JakeSays> ugh. i's going to take 3 days to print this damn thing
<Jack21> printing a menger sponge?
<JakeSays> no. a 160x140x230 box
<JakeSays> open topped box
<JakeSays> it's because cura is just flat out stupid
<Jack21> increase linewidth with less perimeter/shells to save some time (and increase layerheight)
<buZz> or increase wall perimeters to reduce time spent to print infill
<Jack21> iam using bigger lines for top and floor
<Jack21> speed up internal walls
<gbruno> [github] Sploopiii opened issue #4001 (Torus doesnt Render with F6). https://github.com/openscad/openscad/issues/4001
<JakeSays> it keeps adding useless supports
<buZz> JakeSays: why not just disable supports then?
<Jack21> cura gives you a nice summary where it spends the most time
<JakeSays> buZz: i need them
<buZz> oh, i thought you were just printing a open topped box
<JakeSays> buZz: there are hinges that stick out near the top
<buZz> make em only have 45 degree overhangs? then no supports needed ;)
<Jack21> jake still not learned how to use the support blocker ?
<JakeSays> buZz: yeah. i might do that
<Jack21> or set support for only over 85° overhangs
<buZz> hmhm
<buZz> with adequate partcooling you could do 85 yeah
<JakeSays> i dont understand why it won't build supports connected to the model near the top instead of going up from the bed
<buZz> JakeSays: supports connected to the model make the model uglier :)
<JakeSays> Jack21: not sure what you mean by 85deg overhangs
<JakeSays> buZz: yeah but this is a box to hold a mouse. lol.
<buZz> 90 deg = straight angle
<buZz> JakeSays: i dont understand what you mean by that
<JakeSays> yeah i know that
<Jack21> jakesays you can define what angle overhang should be supported ..
<JakeSays> i need supports under these hinge loops
<JakeSays> buZz: i mean that the mouse won't care how ugly his box is
<buZz> -or- reduce the overhang on the 'hinge loops'
<Jack21> because your printer can't bridge?
<buZz> -or- configure the slicer to only support higher overhang than you have
<Jack21> or add custom supports in the design
<buZz> :) yeah
<buZz> meshmixer was/is kinda nice to just 'draw in supports'
<buZz> but windos only afaik , last time i tried it on wine was a HardTime(TM)
<JakeSays> this is what i'm talking about: https://pasteboard.co/1Vo4UF4Y0zAA.png
<Jack21> yeah and it is not supported anymore - last news was that meshmixer is now fusion360
<buZz> you just need the tiniest mods for that to print without supports
<JakeSays> do tell:)
<buZz> 1) 45 degree angles on the straight corners under the loops
<buZz> 2) 45 degree angles on the loops
<JakeSays> i can do that
<Jack21> just print it as it is and remove the first fallen layer later
<buZz> thats also a option :) if it doesnt have to be pretty
<JakeSays> i'd prefer the ugliness be intentional. lol
<buZz> JakeSays: but your mouse doesnt care!
<buZz> :)
<Jack21> use hull() before the differences - Ü
<Jack21> also you are using much higher resolution than needed for this (the slits work with $fn=12)
<Jack21> you are not printing with 5μm layer
<buZz> instead of $fn i usually use $fa and $fs
<veverak> Jack21: well, examples are in 'examples/' folder... anyway yes hex holes are for nuts, never thought about orienting them
<buZz> seems to be a lot nicer
<veverak> Jack21: hmm, I will improve the part "how to use this"
<gbruno> [github] Sploopiii edited issue #4001 ([Solved] Torus doesnt Render with F6). https://github.com/openscad/openscad/issues/4001
<Jack21> veverak and just give default values so you get something without entering anything
<JakeSays> Jack21: how does hull() help with the differences?
<veverak> Jack21: btw: one of the parts in tile.scad is not simple (tile_H) and the 'tile()' itself is decomposed a lot. That happend after I used it extensively on my robot and realised I can't use 'tile()' itsefl directly everytime and sometimes I need just the 'negative' part of it, sometimes just screw positions...
<buZz> hull() makes the shape more uniform
<veverak> Jack21: yeah, I will think abotu defaults
<buZz> you can wrap it around objects and it turns it into one huge blob
<Jack21> jakesays it doesn't it will help with overhangs  and if used after the difference - the diff is gone
<JakeSays> ah ok
la1yv has quit [Read error: Connection reset by peer]
la1yv has joined #openscad
<Jack21> veverak if you use overlap of .1 you will get rid of z-fighting (or use render())
<Jack21> clip in for nuts is also nice feature
<Jack21> veverak also i wouldn't put everything in an extra file .. make ONE example file with the modules in it
<veverak> Jack21: hmm, will have to think about the examples
<veverak> thing is the pics are automatically generated given I use separate files
<veverak> (and that automation is easy)
<veverak> Jack21: 'clip in for nuts' is what?
<Jack21> veverak you can automate this also if everything is in one file
<Jack21> clip in uses a small lip that holds the nut after inserting
<Jack21> or use press marks
<veverak> yeah, but have to think it through
<veverak> Jack21: got any tips for the models with clip in?
<veverak> I understand the idea I think, but wonder what is best way to design it printable
<Jack21> veverak you can add a little sphere above   or add 3 small cylinders around which will be squeezed when a nut is inserted
<veverak> oh god, clang-format broke again for openscad
<veverak> or, I got broken config
<veverak> Jack21: I like the sphere, thanks
mhroncok has joined #openscad
<Jack21> veverak e.g. you can use the animation feature to export all files (without using variables on the command line)
<veverak> looool
<veverak> Jack21: that's nice hack :)
<JakeSays> LOL reduced resolution in some key areas and time to render went from 38 minutes to 6 seconds
<Jack21> Ü
<JakeSays> cura keeps adding supports from the bed to the bottom of the slits
<veverak> hmm
<veverak> well, I eyeballed the sphere dia
<veverak> guess it's time to print something
<JakeSays> ah. increased x/y distance
<JakeSays> it's down to 43 hours
<JakeSays> .2 layer height took it to 37 hours
<JakeSays> Jack21: and no, i still can't get support blockers to work :p
<JakeSays> wtf. when i change layer height from .15 to .2 cura adds back the hinge supports
<veverak> oh fu, I do not use the tile.scad from the repo in the robot's project :facepalm:
<veverak> (and yeah, I did some changes tehre too that I did not propagated into the lib)
<veverak> OT: I can define module multiple times, can I rely on this to be true in longterm?
mhroncok has quit [Ping timeout: 252 seconds]
<teepee> define multiple times?
<teepee> that does not sound like a good idea :)
mhroncok has joined #openscad
Buanderie has quit [Quit: Leaving]
Jack21 has quit [Quit: Client closed]
castaway has quit [Ping timeout: 256 seconds]
Jack21 has joined #openscad
castawayc has quit [Ping timeout: 256 seconds]
castawayc has joined #openscad
castaway has joined #openscad
ochafik has joined #openscad
<Jack21> veverak you can override names yes and you can have different with the same name in different scopes
<Jack21> like you can have 10 loops all using variable i
<gbruno> [github] UBaer21 opened issue #4002 (inconsistent render of transparent object in export png). https://github.com/openscad/openscad/issues/4002
<Scopeuk> interesting, it looks like the viewport is calculating what the transparent cube looks like over the black background and then placing the circle below (hence grey) and the other render options are drawing the cylinder first
<Scopeuk> hmm no its more weird than that, the transparent cube colour shifts quite a bit when moving the view port
<Jack21> also seems one face is not transparent and different shades like light is calculated with different angle
<Scopeuk> and the cylinder is irrelevant to that
<Jack21> hmm so maybe it is just the shading for transparent
<Scopeuk> likely, the preview is all 2d fakery until its rendered (which obviously remove the transparency)
<Jack21> would be nice to influence on the light position .. so you could create 2D images with proper colors
<Scopeuk> perhaps, it is a move away from the core focus however
<Jack21> if i can export 2D geometry i like to have images from 2D geometry not with 3D shading
<Scopeuk> its a fair request and concept.
<Jack21> propably not that simple as 2D are always rendered as 3D in preview
<Jack21> ok managed to print advent № 9 without support and without tipping over .. Warning do not let this lay on the floor unattended - turns out this is a caltrop
<Scopeuk> yeh the 2d subsystem is really focussed arround creating a base for 3d projection
ochafik has quit [Remote host closed the connection]
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
ochafik has joined #openscad
<InPhase> JakeSays: I typically put extreme effort into redesigning parts to not require supports for printing. There are other possible latch designs where there is not a right angle, which might actually end up as stronger parts anyway.
ochafik has quit [Ping timeout: 240 seconds]
ochafik has joined #openscad
ochafik has quit [Ping timeout: 252 seconds]
ochafik has joined #openscad
ochafik has quit [Ping timeout: 240 seconds]
ochafik has joined #openscad
ochafik has quit [Ping timeout: 240 seconds]
ochafik has joined #openscad
TheCoffeMaker has quit [Ping timeout: 256 seconds]
TheCoffeMaker has joined #openscad
ochafik has quit [Ping timeout: 252 seconds]
ochafik has joined #openscad
ochafik has quit [Ping timeout: 240 seconds]
ochafik has joined #openscad
rogeliodh has quit [Quit: The Lounge - https://thelounge.chat]
rogeliodh has joined #openscad
ochafik has quit [Ping timeout: 240 seconds]
ochafik has joined #openscad
<lf94> couldnt find the inspiration to do day 9 last night.
<lf94> I might put it off
<Scopeuk> there does seam to be a dice and dice like object theme going on this year
jkent3 has joined #openscad
jkent3 has quit [Client Quit]
knielsen_ is now known as knielsen
<Scopeuk> ok I printed one of todays shapes too. and it sets off every "curling" anxiety whilst printing due to the shape
<Scopeuk> and Jack21 I concur with the caltrop remark
ochafik has quit [Remote host closed the connection]
ochafik has joined #openscad
noonien has joined #openscad
Jack21 has quit [Quit: Client closed]
Jack21 has joined #openscad
<InPhase> Jack21: If you want to get thousands of downloads on thingiverse, flip tomorrow's Advent Calendar entry upside down, and use it as the base for customizable TPU sandals / flip-flops.
<InPhase> Jack21: Some interest counts are hard to gauge, but that one I can gauge.
<Jack21> hmm curious what we have tomorrow
<Jack21> haha  yeah the surface may work
<Jack21>  or filled with water in the freezer and you have something similar to crushed ice
<Jack21> InPhase just wonder should be possible to get this on both sides if offset by half a cell
<InPhase> Dual purpose! Sandals that also make crushed ice.
<lf94> i'm still scared of day 9
<Jack21> InPhase leaves slightly cheesy aftertaste
<Jack21> lf94 if someone showed this to me - i had no clue how to do this intentionally
<InPhase> Yeah, this one is one of those mystifying representations of sine waves. There is probably a way to re-express it as one function but that's not obvious to me.
<Jack21> and it is only the hull without the intersections
<Jack21> here is another one https://www.prusaprinters.org/prints/88068
<InPhase> Jack21: It's also somewhat of a hack. hull shouldn't really work that way, but it happens to because of how it's implemented.
<InPhase> But, it's a pretty hack. :)
<InPhase> What we should really have is a properly defined points hull.
<lf94> Yeah the hull is what "makes this easy in openscad".
<lf94> I'm not even going to attempt the same. I'm going to have a rounded cube and that other shape and morph between them and call it a day.
<lf94> "that other shape" is the hard part
<lf94> It's like a cone with the point extended
<Jack21> lf94 it is a tetrahedron with bulged faces  and you can generate this with an extruded circle that is asymmetrically scaled
<InPhase> The scaling must be sinusoidal though I think.
<Scopeuk> lf94 I think functionally it might be the overlap of three large spheers
<Scopeuk> 4**
<InPhase> Scopeuk: No, that would have three sharp edges. This is rounded.
<Scopeuk> true there is a soft edge
<InPhase> The circle thing would do it though with [0,1] and [1,0] scaling with a sinusoidal parameter on z.
<InPhase> 180 degrees of cosine plus one over two, I suppose.
<Scopeuk> essentially it reminded me of the reprap Simpson build volume which is what started me off that way but yes thats only a "90%" solution
PovilasCNC has quit [Ping timeout: 240 seconds]
PovilasCNC has joined #openscad
Jack2183 has joined #openscad
Jack21 has quit [Ping timeout: 256 seconds]
Jack2183 is now known as Jack21
jkent3 has joined #openscad
jkent3 has quit [Client Quit]
<gbruno> [github] jcwren opened issue #4003 (WIndows 64-bit 2021.12.07 fails to start due to missing DLL entry point). https://github.com/openscad/openscad/issues/4003
<gbruno> [github] jcwren edited issue #4003 (WIndows 64-bit 2021.12.07 fails to start due to missing DLL entry point). https://github.com/openscad/openscad/issues/4003
<teepee> huh?
ochafik has quit [Remote host closed the connection]
ochafik has joined #openscad
<Jack21> InPhase   regarding pattern  this was made with SCAD https://pasteboard.co/oIXY3QH0w2G2.png
<lf94> Yummy
<Jack21> which is why i think we should have a option for 2D render to have color without 3d shader
<lf94> Anyone know of any research papers for patterns concerning tactile feel?
<Jack21> lf94  there are some about vibration and electricity to simulate the feel of bumpy surface while there is none
<Jack21> the new 3d gloves work with pneumatic pattern to emulate real surface structures
ochafik has quit [Ping timeout: 252 seconds]
<Jack21> apple using controled vibration to simulate a tactile click feeling without real z movement
ur5us has joined #openscad
<Jack21> lf94 what are you looking for - there was one study showing that people with anorexia have reduced tactile abilities
ur5us has quit [Ping timeout: 265 seconds]
ur5us has joined #openscad
<lf94> something that talks about tactile feel of textures
<Jack21> you mean how the sensation is generated?
rogeliodh has quit [Quit: The Lounge - https://thelounge.chat]
ur5us has quit [Ping timeout: 252 seconds]
rogeliodh has joined #openscad
<Jack21> in short you have cells that detect vibration caused by the papillary skin  when moving over a surface https://en.wikipedia.org/wiki/Tactile_corpuscle
Jack21 has quit [Quit: Client closed]
Jack21 has joined #openscad
mhroncok has quit [Quit: Leaving.]
<gbruno> [github] t-paul closed issue #4001 ([Solved] Torus doesnt Render with F6). https://github.com/openscad/openscad/issues/4001
ur5us has joined #openscad
af has quit [Ping timeout: 265 seconds]
ScottJackson[m] has quit [Ping timeout: 260 seconds]
Cadair has quit [Ping timeout: 240 seconds]
t-paul[m] has quit [Ping timeout: 240 seconds]
NoGare[m] has quit [Ping timeout: 245 seconds]
unrust[m] has quit [Ping timeout: 250 seconds]
LeviHarris-Brown has quit [Ping timeout: 268 seconds]
LeviHarris-Brown has joined #openscad
GNUmoon has quit [Ping timeout: 276 seconds]
Cadair has joined #openscad
af has joined #openscad
ScottJackson[m] has joined #openscad
NoGare[m] has joined #openscad
t-paul[m] has joined #openscad
unrust[m] has joined #openscad
GNUmoon has joined #openscad
ur5us has quit [Remote host closed the connection]
ur5us has joined #openscad
Cadair has quit [Write error: Connection reset by peer]
t-paul[m] has quit [Read error: Connection reset by peer]
LeviHarris-Brown has quit [Read error: Connection reset by peer]
unrust[m] has quit [Remote host closed the connection]
ScottJackson[m] has quit [Write error: Connection reset by peer]
NoGare[m] has quit [Read error: Connection reset by peer]
ScottJackson[m] has joined #openscad
little_blossom has quit [Quit: little_blossom]
little_blossom has joined #openscad
ScottJackson[m] has quit [Quit: Client limit exceeded: 20000]
NoGare[m] has joined #openscad
unrust[m] has joined #openscad
t-paul[m] has joined #openscad
Cadair has joined #openscad
LeviHarris-Brown has joined #openscad
Jack21 has quit [Quit: Client closed]
Jack21 has joined #openscad
ur5us has quit [Ping timeout: 250 seconds]
noonien has quit [Quit: The Lounge - https://thelounge.chat]
t-paul[m] has quit [Quit: Client limit exceeded: 20000]
<JakeSays> ugh. i think i've got a bad spool of petg
ScottJackson[m] has joined #openscad
t-paul[m] has joined #openscad
Jack21 has quit [Quit: Client closed]
Jack21 has joined #openscad
<Jack21> .. aslong it isn't the filament - Ü
<Jack21> these paper spools can cause issues - so i heard
<JakeSays> LOL
<JakeSays> its the filament:p
<Jack21> i am sleepy but i will open my door .. although i knew what is in there ..
<Jack21> inconsistent diameter or fixable like wet
<JakeSays> i'm sure some of it is my printer calibration, but the rest has got to be the filament. i'll take a pic.