<Scopeuk>
there are ways of converting a mesh to a step, but you just end up with a mesh inside the step file. people don't like those either
<Scopeuk>
perfect world step export would use the base primitives to form the model in the step and probably name assemblies based on modules
<J23k>
can't the CSG tree be converted? sure not all circles are circles
<J23k>
FreeCAD did convert circle(); and cylinder();
guso78k has joined #openscad
<guso78k>
probably need a step file renderer ? :)
<guso78k>
and subclass Renderer() ...
<J23k>
As STEP is industry standard - such an export would enable SCAD to be included into the workflow of companies ( in theory )
<J23k>
and slicer (bamboo prusa) can read STEP too
<J23k>
CURA has a plugin for CAD support The plugin is approximately 500MB in size,
<Scopeuk>
I think the "fun" comes in with things like minkowski
mmu_man has joined #openscad
<J23k>
the first rule of the minkowski club - Don't use minkowski
<Scopeuk>
it would be fun to scad something up, kick it out to a "conventional" parametric cad tool edit and come back. but that would REALLY pushing the import/export
J23k has quit [Quit: Client closed]
J23k has joined #openscad
J23k has quit [Quit: Client closed]
J23k has joined #openscad
J23k has quit [Quit: Client closed]
J23k has joined #openscad
teepee_ has joined #openscad
teepee has quit [Ping timeout: 246 seconds]
teepee_ is now known as teepee
guso78k has quit [Quit: Client closed]
L29Ah has joined #openscad
<teepee>
J23k: yep, csg tree can be converted to step, freecad proves that, it would have been so much nicer if that would be on our side
<teepee>
only problem is that STEP is not a file format but an operating system or so
<J23k>
oh
<teepee>
I tried "exporting" an "empty" file some time ago, not with OCCT but with StepCode and that was 14k text
<teepee>
like introducing what geometry is, what coordinates are and ponts and whatnot
<teepee>
also there's not "STEP" there like hundreds? of specifications for different use cases and I don't really looked into it how that applies
<J23k>
ok but for a 2mb file a 14k overhead doesn.t seem to be a big issue
<teepee>
so in therory it's not *that* complicated
<teepee>
true
<teepee>
just saying it meant to do *everything* making it hugely complicated
<J23k>
I guess we need a super simple "just circle and straight lines"
<teepee>
plus at the time when I was looking there was 2 different libraries in debian that were incompatible
<teepee>
oh, it can mostly represent the CSG tree is my understanding
<teepee>
so if you drop hull() and minkowski() it *should* translate mostly 1:1
<teepee>
well, so I heard :)
<Scopeuk>
yeh step has huge numbers of appendix depending on what you are currently doing. I don't believe it to be internally incompatible but it is far from trivial
<J23k>
so we get design warnings for hull that this will not be STEP compatible - Ü
<teepee>
wiki criticism: "the format is not storage-efficient. For example assigning an RGB color code to an edge requires at least 6 other entities, and specifying a transformation requires at least 5 additional entities (PLANE, AXIS2_PLACEMENT_3D, a CARTESIAN_POINT, and 2 DIRECTION entities)
<dTal>
Seems to me that rendering scad to STEP is always going to be second class and as such should probably be a separate project
<teepee>
again "in theory" IGES might be a better choice, but the existing library list is pretty thin (I know about just one) so not ideal from the technical aspect
<teepee>
someone actually started that but I never could get it to work at the time due to that multi-library issues
<Scopeuk>
WARNING: The conversion is not deterministic,
<J23k>
"needs several times"
<J23k>
is this like AI .. you get 9 results and choose the best fitting?
<juri_>
standardizing scad is a long term goal of mine... but i have too many long term goals.
aiyion1 has quit [Ping timeout: 246 seconds]
aiyion1 has joined #openscad
J23k has quit [Quit: Client closed]
J23k has joined #openscad
mmu_man has quit [Ping timeout: 260 seconds]
mmu_man has joined #openscad
snaked has quit [Quit: Leaving]
guso78k has quit [Quit: Client closed]
mmu_man has quit [Ping timeout: 246 seconds]
<teepee>
J23k: yes, that comment is a bit concerning
aiyion1 has quit [Ping timeout: 246 seconds]
mmu_man has joined #openscad
<Scopeuk>
"it works all the time some of the time" screams invalid memory usage to me but I've not gone over their code, it could be a race condition of some description I guess.