TheCoffeMaker_ has quit [Ping timeout: 252 seconds]
TheCoffeMaker has joined #openscad
killjoy has quit [Ping timeout: 240 seconds]
killjoy has joined #openscad
killjoy has joined #openscad
killjoy has quit [Changing host]
arogora has joined #openscad
killjoy has quit [Ping timeout: 272 seconds]
arogora has quit [Ping timeout: 255 seconds]
killjoy has joined #openscad
killjoy has joined #openscad
killjoy has quit [Changing host]
arogora has joined #openscad
killjoy has quit [Ping timeout: 272 seconds]
arogora is now known as killjoy
killjoy has quit [Changing host]
killjoy has joined #openscad
califax has quit [Remote host closed the connection]
califax has joined #openscad
mmu_man has joined #openscad
L29Ah has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 252 seconds]
teepee_ is now known as teepee
J23k53 has quit [Quit: Client closed]
J23k53 has joined #openscad
cart_ has joined #openscad
misterfish has quit [Ping timeout: 260 seconds]
<lf94>
guso78: InPhase: peeps[work]: https://github.com/lf94/summonscript - WIP, but that's what's coming. I haven't shared it with anyone else, I'm trying my best to hold back before Im done a demo and some examples :D.
<lf94>
I'd love to see you guys blow everyone out of the water
J23k53 has quit [Quit: Client closed]
<dTal>
"No macros allowed for feature creation. "
J23k53 has joined #openscad
L29Ah has joined #openscad
<dTal>
Depending on whether they define that as pre-existing, or at all, that might disqualify OpenSCAD altogether :p
<dTal>
I dunno if it's possible to "blow everyone out of the water" - the top time is 2:53
<dTal>
"You must CLEARLY show the correct mass of each model, before moving on to the next model. The mass must be shown in the CORRECT UNITS. The mass must be shown with the the CORRECT PRECISION (or with a MORE PRECISE number of significant digits)." -- another difficulty
<lf94>
For almost all those models aspects, CSG would be super fast
<lf94>
the 2nd model is just a trace + revolve
<lf94>
The hardest one looks like the first one
<lf94>
with the fillet against the circle
<lf94>
dTal: someone used build123d which is a code CAD system too
<dTal>
You think you could type in the entire model for one of those and verify its mass in under a minute?
<dTal>
Verifying the mass is gonna require rendering in very high res, exporting the STL, computing the volume with a third party program, and converting the units to the desired units with another third party program
invultri has quit [Quit: leaving]
<lf94>
> exporting STL < well I guess that's an openscad problem
<lf94>
Rendering in high res shouldnt be too much of a problem
<lf94>
considering these are all pretty basic
<lf94>
> in under a minute < not the first one unfortunately. that one fillet looks like a pain for any non-brep kernel
<lf94>
I think you're right that the guy in first is probably optimal in operations
<lf94>
it basically comes down to how fast can you click or type a sketch
<lf94>
Still these guys are professionals vs us non professionals
Guest17 has joined #openscad
Guest17 has quit [Client Quit]
califax has quit [Ping timeout: 252 seconds]
califax_ has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 252 seconds]
teepee_ is now known as teepee
califax_ is now known as califax
<dTal>
The thing is that a professional CAD system will alow you to enter most of those dimensions as constraints, whereas with OpenSCAD you need to calculate the constraints yourself
<dTal>
So for example that first part, you have those side flanges which are made primarily of a circle profile and a rectangle profile, filleted together. You know the radius of the circle and the width and height of the rectangle, those are given. But how to translate the rectangle? You need to do math: you're given the width of each component and the total width, so you have to work out the offset.
<InPhase>
lf94: If timing is what those come down to, and you can practice ahead of time, all we need is a hackertyper style setup to win all of those with OpenSCAD. :)
<InPhase>
Randomly mash the keyboard, let the text stream into the editor, and boom, done.
<lf94>
haha
<lf94>
I mean we do want to be as fair as possible
<InPhase>
Is everyone else? :)
<lf94>
The most fair is probably you must not look at the model before doing the test
<teepee>
I think it's a pointless exercise
<lf94>
I'm not sure but I don't really care if they're not, I care how proficient we are
<lf94>
It's exercise; it all counts for something
<lf94>
It certainly means something if an OpenSCAD programmer can compete with a professional CAD user
<InPhase>
lf94: Looks like everyone else is pre-practicing the clicks, and then it's an exercise in how fast you can click through the menus and such.
<lf94>
Yeah that's lame
<InPhase>
lf94: The equivalent for us is pre-writing the script and making it short and compact, and then typing it in quickly.
<teepee>
why not copy&paste?
<lf94>
Yeah yeah. I would much prefer blind tests, where no one knows what's coming
<InPhase>
This is of course a silly way to run a competition.
<InPhase>
Yeah, it would have to be everyone legitimately encountering the problem for the first time.
<InPhase>
The real test is how quickly can you devise a solution given a particular system.
<lf94>
yee
<InPhase>
And honestly I think we'd be pretty competitive on a lot of those shapes.
<lf94>
Mhm, my thoughts too
<lf94>
I really want a small group of code CAD users to get together and do a professional project
mmu_man has quit [Ping timeout: 255 seconds]
<lf94>
to test the strength and weaknesses of code CAD
<lf94>
Like build a bridge or something
<InPhase>
#2 and #3 are trivial. I bet that takes me very little time. The longest amount of time would be hunting for the numbers on the spec sheets, since I haven't used a spec sheet like that for anything serious in like 15 years.
<InPhase>
#1 in their example would have a slight lag on getting that inward curve lined up right, since that was clearly designed to be specified in a gui cad system. Still doable, just it's not clearly specified what are the edges of that to make it smoothly differentiable if the inner radius is 0.75 and the outer radius is 1.5.
<InPhase>
Even looking at it I'm not entirely sure it has one answer from the values written there.
<InPhase>
One might have to introduce an unspecified 45-degree switchover assumption?
<dTal>
It looks pretty clearly specified to me
<InPhase>
What constrains where the 1.5 curve becomes the 0.75 curve?
<dTal>
math
<InPhase>
On what values?
<dTal>
the two curves must be tangent at the point of intersection
<InPhase>
Nothing locks down the angle of sweep or the coordinates or the dimensions of either curve to the crossover point.
<dTal>
If you imagined that fillet were not there, and you had a circle of the correct radius, and you nestled it in the corner, there would be only one place it could go
<InPhase>
Oh, yeah, I see it. The 1.5 to the 2.625 and the tangent constraint.
<dTal>
the challenge wouldn't be making the openscad file, but golfing it
<InPhase>
These sorts of shapes would be pretty easy to golf once in place.
<InPhase>
Audio commentary "And then we type in the text".
<dTal>
I'm thinking it's gonna be a heck of a typing challenge
mmu_man has joined #openscad
<dTal>
I'm on 22 lines just for that profile
J23k53 has quit [Quit: Client closed]
J23k53 has joined #openscad
<dTal>
if you could alias things like "translate" and "square" that would help