<ccox>
teepee - yes, I know. I was just looking at the code.
<ccox>
and the question of how to handle the settings is independent of how to handle the code.
<teepee>
I suppose there's 2 options for the settings, the preferences window and the color scheme
<teepee>
it might be possible to cover both with the preferences allowing to overwrite the color scheme
<ccox>
it has to go in preferences one way or another. But do lights belong as part of the render color scheme, or as a separate list of light schemes?
<teepee>
I would not see them as separate scheme to select
<teepee>
having one override in the preferences if the user decides to use that seems enough
<teepee>
if that ever gets requested a number of times, it should be possible to extend
GNUmoon2 has quit [Remote host closed the connection]
GNUmoon has quit [Remote host closed the connection]
GNUmoon has joined #openscad
<dalias>
does csg computation not reuse subtrees?
<dalias>
i have a complex difference that's repeated 40 times in rotation, and it should just compute the surface once and duplicate it, but it seems to be doing some monster computation instead :/
<dalias>
uhg never mind, it has gaps because of numrical imprecision that make it hellishly slow
califax- has joined #openscad
califax has quit [Ping timeout: 276 seconds]
califax- is now known as califax
LordOfBikes has quit [Ping timeout: 256 seconds]
<teepee>
dalias: it does if the effective expanded code is identical
<teepee>
so it does not work when passing things like rotation angles deep into the hierarchy
LordOfBikes has joined #openscad
<dalias>
nothing's passed deep in. it's literally for .... rotate(i*k) constant()
<dalias>
but i think the problem was just the complex almost-touching surfaces between each instance
<dalias>
because of numerical instability
<ndnihil>
what is "roof"?
<InPhase>
dalias: Yeah, that'll definitely slow things down substantially.
<InPhase>
Reuse tends to be pretty good though. I like to throw in module boundaries where I expect reuse to happen to be explicit about it.
<dalias>
*sigh* don't remind me roof still isn't merged :( :( :(
<InPhase>
Looks like fast-csg is the top three open PRs, and the top three open issues. :) That's fast.
zauberfisch has quit []
J2272 has joined #openscad
<ndnihil>
so it's only missing the closing paren and url from pushes...
J22 has quit [Ping timeout: 256 seconds]
<teepee>
InPhase: top three?
<ndnihil>
ok, might be fixed
<ndnihil>
we'll see on next push
teepee has quit [Remote host closed the connection]
teepee has joined #openscad
<InPhase>
teepee: Most recent three, so top listed.
zauberfisch has joined #openscad
<teepee>
the PRs are draft, so they'll stay there for a bit :)
<InPhase>
Mostly I just noticed a whole lot of fast-csg emails streaming in, so I went to check the issue and PR lists on github for a more orderly presentation. Seems to be getting tested more rapidly than most things.
<auctus>
buZz: re: game console; im working on the cartridge reader mechanism, reader + SBC + case = console hardware; im writing software too, bare necessities: dashboard UI + game launcher
<auctus>
i'll get ya a pic later when im back in my garage
<auctus>
basically the cartridge reader is a series of switches that i print, spring-loaded; so by default they're all closed, cartridge has pins on it that push the switches open, so the cartridge has like, 10 teeth say, that gives me room to have 1000 games
<auctus>
then i can 3d-print cartridges, which are really just keys to load games
Guest40 has quit [Quit: Client closed]
<crazy_imp>
auctus: what about fork light barriers? less complex to integrate and you only need sone holes in your cards :)
lastrodamo has joined #openscad
<buZz>
auctus: ahhhh like some of those children record players :)
Junxter has quit [Read error: Connection reset by peer]
arebil has joined #openscad
ToAruShiroiNeko has quit [Changing host]
ToAruShiroiNeko has joined #openscad
<OlivierChafik[m]>
I think I've finally understood a simple way to do the post-corefinement remeshing. No need to call any of the expensive remeshing functions, just need to collect patches of coplanar faces (which we know because we tracked their split throughout the corefinement visitation), then delete them all and replace them with polygons (valid in a surface mesh) and just call the triangulate_mesh helper. Trying to code this now :-D
ABSHK has left #openscad [Leaving]
Administrator_ has joined #openscad
ABSHK has joined #openscad
ToAruShiroiNeko has quit []
ToAruShiroiNeko has joined #openscad
snaked has quit [Remote host closed the connection]
snaked has joined #openscad
felix_5 has joined #openscad
arebil has quit [Quit: My keyboard has gone to sleep. ZZZzzz…]
stefanct has quit [Ping timeout: 268 seconds]
stefanct has joined #openscad
ferdna has quit [Quit: Leaving]
Administrator_ has quit [Ping timeout: 250 seconds]
linext has quit [Read error: Connection reset by peer]
linext has joined #openscad
linext has quit [Read error: Connection reset by peer]
<J2272>
`roof()` is not working well ..:( have fonts and some letter dont't work .. when using an offset .. how can that be?
Netfool has quit [Ping timeout: 256 seconds]
felix_5 has joined #openscad
<felix_5>
he(g8) höö(c2)
<felix_5>
Is there a good way of effectively overload a module to eater accept named parameter or a list of all parameter? I am new to scad and don't know if it is even possible.
<teepee>
good way, no
<teepee>
but you can check the first parameter via is_list() and unpack that
<J2272>
roof() doesn't like minkowski either .. maybe that font is f'up
<teepee>
even with a small offset?
<J2272>
neither pos nor negative .. and also minkowski is not working
<J2272>
negative already gives something like "sceleton calculation error"
<J2272>
the other show fine in preview but result in "nef .. " when render
<J2272>
happens only with n and t so i assume these letter are broken and offset doesn't fix it ( which is not clear to me )
GNUmoon has quit [Ping timeout: 276 seconds]
<J2272>
now i have "OMIA VICI AMOR" :(
<J2272>
which probably inverse that quote to Everything wins over love