00:01 <joseph_> kintel: Right, so I've been thinking that one of the earliest targets of refactoring is unifying the rendering logic between OpenCSGRenderer and CGALRenderer. Helpfully, many of my modifications to CGALRenderer last summer are mostly copied from what was already in OpenCSGRenderer. But it seems to me that I haven't yet touched the code that is actually causing the artifacts: the fixed-function drawing of OpenCSG primitives. I can

00:01 <joseph_> start searching for where that happens, unless you know off the top of your head where it's located

00:32 <peeps[zen]> guso78: that image is manifold as long as the front face has 5 sides (2 colinear at the top edge). every edge consists of two half-edges, each half-edge belongs to exactly one polygonal face

00:40 <peeps[zen]> that's valid AFAIK in CGAL, assuming all the 5 points on the front face are coplanar. i think Manifold library requires triangulation, which would require J23k4's edited version of the image.

00:41 <peeps[zen]> but the necessary triangulation should happen behind the scenes as long as it was valid as previously described

00:42 <peeps[zen]> BTW, J23k4 all your links add an underscore that breaks things and requires manually editing url for me

00:43 <peeps[zen]> i.e. i click your link and browser goes to https://imgpile.com/i/9f6zwk%C2%A0

00:43 <peeps[zen]> so I have to remove the percentage garbage

00:44 <peeps[zen]> hexchat irc client here

02:07 <kintel> joseph_ I started porting stuff over a month or two ago: https://github.com/kintel/OpenCSG/tree/glsl https://github.com/kintel/OpenCSG/tree/opengl2

02:07 <kintel> See also https://github.com/openscad/openscad/pull/4596

02:08 <kintel> My code only touches the SCS algorithm; I didn't yet look into Goldfeather

02:11 <kintel> Btw. in terms of OpenGL modernization, one approach to tease out all the call sites could be to start with https://github.com/openscad/openscad/pull/4582 and request an OpenGL 3 context without compatibility profile. The EGL code isn't finished, but it should work well on macOS

02:24 <joseph_> Thanks for the advice. I'll investigate

05:04 <J23k4> peeps[zen]  regarding the link - i am using https://web.libera.chat/?channel=#openscad  the community link from openscad.org and for me those links look fine  and also in the logs https://libera.irclog.whitequark.org/openscad/2023-06-02 they work and look as intended.  Is it possible that your client made the conversion?

05:05 <peeps[zen]> J23k4: links from other users (like guso78's original image) work without editing

05:05 <J23k4> aren't they all use matrix?

05:07 <J23k4> still wondering why the log did't show this too .. maybe browser related

05:08 <JackLink> link test https://openscad.org/cheatsheet/snapshot.html

05:09 <J23k4> looks fine for me

05:10 <peeps[zen]> that one is fine. idk, your imgpile link from earlier was messed up though, and i noticed it a couple other times hadn't bothered mentioning it until now.

05:10 <J23k4> https://openscad.org/cheatsheet/snapshot.html

05:11 <J23k4> what is with this one

05:11 <peeps[zen]> no problem there

05:12 <peeps[zen]> apparently its an nbsp, not an underscore that I said earlier

05:12 <J23k4> https://imgpile.com/i/9f6zwk  this was the link

05:12 <J23k4> https://imgpile.com/i/9f6zwk

05:13 <peeps[zen]> https://www.w3schools.com/tags/ref_urlencode.ASP Ctrl-f %C2%A0

05:13 <peeps[zen]> J23k4: second one is good, first one is bad

05:14 <peeps[zen]> someone was complaining about nbsp in here the other day, dunno if that's related. i have no idea what these weird web clients do

05:14 <J23k4> i added a double space on the second

05:15 <peeps[zen]> ah well there you go, i guess that's the solution

05:15 <peeps[zen]> might be a bug to file with the web client

05:16 <J23k4> https://example.org/badlink https://example.org/goodlink

05:16 <peeps[zen]> those are both good *shrug*

05:16 <J23k4> https://example.org/badlink

05:16 <J23k4> no space

05:17 <peeps[zen]> that's fine too. i think it maybe specifically a link followed by non-link text

05:17 <J23k4> https://example.org/badlink

05:18 <J23k4> https://example.org/badlink A

05:18 <peeps[zen]> hmm, they're all fine now

05:18 <J23k4> ok i am out of ideas

05:19 <J23k4> https://imgpile.com/i/9f6zwk a

05:22 <J23k4> https://imgpile.com/i/9f6zwk a

05:22 <J23k4> https://imgpile.com/i/9f6zwk  a

05:26 <peeps[zen]> bad, good, bad

05:27 <J23k4> wow

05:27 <peeps[zen]> first one includes nbsp and the "a", last one has just a trailing nbsp

05:28 <peeps[zen]> as part of the link

05:28 <guso78> haha, impresssed that such a simple pic get so popular :') :')

05:28 <J23k4> this seems to be a bit random

05:29 <J23k4> https://example.org/badlink

05:29 <peeps[zen]> good

05:29 <J23k4> https://imgpile.com/i/9f6zwk a

05:29 <peeps[zen]> bad

05:30 <J23k4> ok this was a copy of the indicated bad one

05:31 <J23k4> https://imgpile.com/i/9f6zwk plain text copy

05:35 <peeps[zen]> good

05:39 <J23k4> hm so i assume something get into the copy when taking it from the browser bar

05:40 <J23k4> https://example.org/

05:40 <J23k4> https://example.org/ a

05:40 <J23k4> https://example.org a

05:40 <J23k4> https://example.org  a

05:42 <J23k4> https://example.org/ plain  - will use that in future .. please let me know when you spot any inconvenience

19:48 <gbruno> [github] burunduk3 opened issue #4658 (CSG export unexpectedly changes directory) https://github.com/openscad/openscad/issues/4658

20:12 <gbruno> [github] burunduk3 closed issue #4658 (CSG export unexpectedly changes directory) https://github.com/openscad/openscad/issues/4658