09:41 <Guest61> Hi All, I would like to create a half circle like shape to be used with BOSL2 offset_sweep. Here's what I came up so far:

09:41 <Guest61>     include <BOSL2/std.scad>

09:41 <Guest61>     include <BOSL2/rounding.scad>

09:41 <Guest61>     s = right_half(circle(d=15, $fn=30), x=1);

09:41 <Guest61>     polygon(round_corners(s, r=1));

09:41 <Guest61> I would like to have it smoother, but increasing $fn yields an error that I can't really interpret. I'm not that familiar with OpenSCAD. How can I make the shape more smooth?

09:54 <Scopeuk> where are you adding the $fn?

09:55 <Scopeuk> the current one is the circle number of facets rather than the facet count of the extrude

09:55 <Scopeuk> it looks like round_corners also takes a $fn argument

09:55 <Scopeuk> polygon(round_corners(s, r=1),$fn=30);

09:56 <Scopeuk> I'm not familiar with BOSL to go much further

09:56 <Scopeuk> other people may be able to jump in later

10:01 <Guest61> The original code had $fn lifted out, I just moved it to the circle to make it 1 line shorter:

10:01 <Guest61>     $fn=31; // fails!

10:01 <Guest61>     s = right_half(circle(d=15), x=1);

10:01 <Guest61>     polygon(round_corners(s, r=1));

10:02 <Guest61> I would like to get the circle smoother, but your idea helps with the section where the rounding kicks in.

10:08 <Scopeuk> I suspect 31 is causing issues due to not being even, that however is an informed guess rather than based on anything solid

10:10 <Scopeuk> there are some notes in the bosl lib https://github.com/BelfrySCAD/BOSL2/blob/17f0fe8428d0d730505122b128f20ee43d0f8a7c/rounding.scad#L136 that go over things

10:10 <Scopeuk> this includes some scenarios that cause errors and explaining you can add verbose=true to get some more help with errors

10:17 <Guest61> Yeah, I got a list of "scale factors", but I don't know what to do with them, and I don't understand why I need put scale factors from error messages into the code depending on the value of $fn. It sounds workaroundish, and I thought there's a better way to achieve the same result.

10:42 <Scopeuk> in this case Bosl is trying to do some specific warping of the part to achieve the effect, under some circumstances the way the numbers work out the maths for some input combinations breaks down . those numbers are designed to move the numbers in such a way as you get the same result but without hitting the maths issues. I don't know BOSL well

10:42 <Scopeuk> enough to go into any more detail on this particular case

10:44 <Guest61> thanks, I'll look into it!

17:52 <teepee> oh, my, the technoking forgot to also mess with the favicon, what a BS

17:52 <teepee> can we just go with Mastodon now?

18:03 <teepee> RIP ChanServ

18:21 <Scopeuk> It is risen!

18:21 <J23k89> that was fast

18:24 <InPhase> But is it like, done for real?

18:24 <teepee> I have not seen any announcement yet

18:24 <InPhase> Yeah. I figured, wait a little for one.

18:25 <teepee> actually that might be confirmation that all is done: <tomaw> I did most of it before the reboot

19:06 <teepee> also they finally found the favicon :(

19:18 <teepee> ahh, official notification, nice

23:15 <teepee> that looks cool https://www.printables.com/model/527993-another-parametric-matrix-keypad