Behdad Esfahbod
Behdad Esfahbod
Nice. Can you modify it to return all solutions? There may be at least two or even three points at the same distance. Or maybe this is fine to have...
> interesting, thanks for sharing. Could you also provide a link or citation to where you discovered this (paper or source code etc.)? It's a numerical brute-force. It samples n...
I agree it probably be very hard for platforms to embrace such API. Leaving it around for a while to see if I can get to it.
What would an API for this look like possibly?
I'm more open to just returning `no layout change` / `possible layout change` per variation axis. That's more feasible. And at a face level. That would look into all of...
> Would this `no-layout-change` also cover side-effects of one axis triggering others through `AVAR` 2? You just pointed out why this might not be trivial :-). Still doable, but more...
Why do you need to instantiate `STAT` separately? Try: ``` from fontTools.varLib.instancer import AxisRange instantiateSTAT(font, {"wght":AxisRange(500, 900), "ital":AxisRange(0,16), "mono":AxisRange(50,50)}) ```
If we do this we should also use it in the `CFFToCFF2` module.
Thanks for the report. I'll study it in detail. But so you know, `hb_font_t` is very cheap to create, if you can get yourself to just do that. You can...