Open breznak opened 6 years ago
Don't make a gamma correction protocol, make colour space manipulation protocol. You probably stand a greater chance of having it accepted in Wayland if you do (your chance is virtually zero if you make a gamma correction protocol), and it is what Redshift ought to be using anyways. The protocol should support multiple concurrent manipulations, and since matrix multiplication is non-commutative, that means the protocol should use order hints. Basically, you want to do https://github.com/maandree/coopgammad, but with colour space matrices instead of gamma ramps. You may even what to have a protocol that supports both.
For reference https://bugs.freedesktop.org/show_bug.cgi?id=97434 asks for this be be fixed, and the response is (paraphrase) "file bugs for all the compositors" because the respondent thinks it requires modification of screen content, and so is out of scope for Wayland. Perhaps strictly true, but Wayland should guard against alterations to what screen content means, which has some independence from what it looks like-- generally difficult to do, though. Can altering colour temperature of the entirety of all displays be exploited? I guess that's what maandree is alluding to.
...as those cannot be supported currently by redshift, and instead develop their own duplicate solutions. https://github.com/jonls/redshift/pull/568#issuecomment-356676362