tynet-lichat/shirakumo - IRC Chatlog
Search
17:31:22
|3b|
though i suppose if they did, using it might actually be appropriate for some uses of those functions
17:32:05
|3b|
and actually i guess user should be able to supply a 'previous character' anyway, so kerning is correct across things like color changes that end up as 2 calls
20:28:46
|3b|
ACTION was assuming % was more internal, so call chain went other way when just looking at diff and remembering how i patched it before
20:29:30
shinmera
Yea I wasn't sure what to call the other kerning function that goes straight through a table.
20:29:49
|3b|
ACTION has some more optimizations for the parts i use once it is done, i think, but seems to work as is
20:49:55
shinmera
From my side it's ready to merge, unless you also wanted me to change the other plist parts into structs like I mentioned.
20:50:19
shinmera
Also I'd do a squash merge in any case cause my commit style isn't exactly pristine, if that matters to you :y
20:51:20
|3b|
ok, will try to get to that soon if clothespin (the other user i know or) doesn't have any comments
21:31:35
|3b|
ok, most of the remaining known errors gone (last one that is actually visible is arguably due to bad font anyway, messy tiny details that sdf can't reproduce and that are hard to remove)
21:32:16
|3b|
there are still a bunch of glyphs that look like zpb-ttf didn't parse them correctly that i need to look at later
21:34:38
|3b|
ACTION needs to add another test where i compare result of rendering through generated SDF against rasterizing through some non-lisp thing
21:35:54
|3b|
though actually that sounds annoying... probably hard to get things lined up right, and convince it not to do any hinting or whatever