<lockywolf>
What do people to for texturing openscad designs?
<teepee>
my guess: nothing
<teepee>
in the sense it's not a common thing done
<lockywolf>
well, I get it that 3d printers don't care if the design is textured, but I suspect it would be really nice for previewing
<teepee>
with all the graphics updates lately that might be something for the future now, but at this point there's no support for that
<lockywolf>
well, I didn't expect openscad to be able to do it
<teepee>
you can eport to povray now and do full raytracing
<teepee>
*export
<lockywolf>
but I thought that there is some idiomatic pipeline, as in "design in openscad" -> export to software X for texturing -> pass to yafaray for rendering
<teepee>
I was hoping for yafaray, but over the last 10 years it sadly looked like broken, abandoned, or something else in varying degrees
<teepee>
hum, and now the domain is for sale
<lockywolf>
well, the actual renderer is not that important
<lockywolf>
the first link in Google is "meshmixer", but it is Autodesk, so I thought there should be something open source
<lockywolf>
perhaps Blender?
J25k86 has joined #openscad
J25k1 has quit [Ping timeout: 240 seconds]
mmu_man has quit [Ping timeout: 268 seconds]
guso78k has quit [Ping timeout: 240 seconds]
fling has quit [Remote host closed the connection]
fling has joined #openscad
foul_owl has quit [Ping timeout: 268 seconds]
zarac has joined #openscad
drfff has joined #openscad
lastrodamo has joined #openscad
foul_owl has joined #openscad
JakeSays has quit [Remote host closed the connection]
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
kintel has joined #openscad
<kintel>
Right, linear extrude too. That does sounds like an interesting source of corner cases
<guso78k>
I am just watching the comparison between OpenSCAD and CadQuery in the OpenSCAD mailing list. . IMHO none of these facts is true anymore ...
<guso78k>
what would also be cool: Sanjeev has written a surface_extrude sample with a waved surface(built from many triangles). using that as a source for linear_extrude would be cool ...
<teepee>
while those examples are always extemely impressive, every time I've asked for a general solution, that was not there
<teepee>
giving users features that are only working under certain not explained conditions is not going to work well
<scrameta>
I guess there is a fair bit of room in the definition of the language features that is 'undefined behaviour'.
<guso78k>
understand your perspective and i completley agree to the need of stability. But only admitting things which are 200% correct gates others learn about new ideas, visions. they do not have a chance to think about it and come up with solutions. I was thinking, that experimental feature was exactly for that purpose ...
<guso78k>
scrameta, your skin rework is absolutely cool. I like most the ability to place 2D stuff arbritrary in space, still i have fixed few display bugs
<guso78k>
and yes, i think that the skin stuff is prone to self-interectoin, but people which use it are well aware. I dont care at all ...
<scrameta>
I can't take credit for the core of this, it was an old PR I updated. My only contribution is the different number of vertices/alignment stuff.
<scrameta>
There was some protection in the code for slice->slice intersection, but not on the overall level.
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<scrameta>
The python stuff is neat.
<scrameta>
I think it needs a feature to extract the coordinates from any child
<scrameta>
If that doesn't already exist
<guso78k>
scrameta, since 2 hours, skin can be called from python. did you get it compiled
<guso78k>
scrameta, you can use mesh() to extract coordinates from any solid ...
<scrameta>
So it can be made from primitives + modules, then extracted to coordinates and then put back into a polyhedron
<scrameta>
Ah cool
<guso78k>
yes, exactly and you can make crazy nonlinear coordinate transformartions in between ...
<scrameta>
Yeah I got that compiled and it's working.
<guso78k>
Excellent!
<scrameta>
I was expecting it to take a list though rather than a,b,c,...n. Though that works well I was just thinking union takes a list.
<scrameta>
or maybe it doesn't I've not used the python stuff much yet
<scrameta>
Love the mesh function!
<guso78k>
difference([a,b],[c,d],e,f) a,b are positive, c,d,e,f are negative
<guso78k>
basicaly a list as an argument makes union implicitely
<guso78k>
union([a,b]) will also work, but its way easier to write: a|b
<scrameta>
So if I call mesh can I do the inverse? I was thinking I could do polyhedron(mesh(some_shape)) but its not that simple.
<guso78k>
yeah, you can do:
<guso78k>
vertices, triangles = mysolid.mesh()
<guso78k>
then make fancy stuff on the vertices
<guso78k>
then polyhedron(new_vertices, triangles)
<scrameta>
Nice
<guso78k>
this only works to some extent as there is nothhing which would break a long segment into many small segments(which would be needed to make a line round)
<scrameta>
Seems a supporting function could allow that fairly easily... somehow
<guso78k>
yeah, a function, which needs to do that is wrap().. It wraps a thin long solid around a cylinder with radiud given
TheAssassin has joined #openscad
TheAssass1n has quit [Ping timeout: 264 seconds]
teepee_ has joined #openscad
teepee has quit [Ping timeout: 264 seconds]
teepee_ is now known as teepee
<guso78k>
teepee, the VMF functions from BOSL2 nicely demonstrate, how triangles coming from completely differenc sources can form a water-tight solid. Sanjeev could do that for his waved surface, too.
<teepee>
the external code is cool, but it only helps *us* if someone adds the feature in openscad for all internal stuff, or explains the algorithms in a way someone else can do the coding
<guso78k>
the PR from scrameta was in *really* good shape IMHO, but only after merging i was best able to understand it and fixing a glitch in Render(F6) of 2D shapes
TheCoffeMaker has quit [Ping timeout: 252 seconds]
teepee_ has joined #openscad
teepee has quit [Ping timeout: 264 seconds]
teepee_ is now known as teepee
<teepee>
meh, connection is extra unstable today :(
<scrameta>
This rendering bug fix - something I should cherry pick back out to the PR to openscad?
<teepee>
I have not followed the details, but if it's a separate fix, that seems like a good idea