Re: [PATCH] New function: color-blend

[Eli Zaretskii]

> From: Joseph Turner <joseph@breatheoutbreathe.in>
> Cc: emacs-devel@gnu.org
> Date: Sat, 02 Nov 2024 11:00:46 -0700
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> >> From: Joseph Turner <joseph@breatheoutbreathe.in>
> >> Date: Fri, 01 Nov 2024 22:29:41 -0700
> >> 
> >> Here's `color-blend', a function to blend two colors into one.  The
> >> patchset also replaces `vtable--color-blend' with `color-blend'.
> >> 
> >> Examples of this function being implemented separately in the wild:
> >> 
> >> https://github.com/alphapapa/prism.el/blob/2fa8eb5a9ca62a548d33befef4517e5d0266eb28/prism.el#L1044
> >> https://oremacs.com/2015/04/28/blending-faces/
> >
> > Thanks.
> >
> >> +---
> >> +** New function 'color-blend'.
> >> +This function blends two RGB lists into one.
> >
> > This should explain what the function does, in terms that are clear
> > even to those who are not experts in this area.  AFAIU, "color
> > blending" has at least two different meanings.
> >
> >> +(defun color-blend (a b &optional alpha)
> >> +  "Blend the two colors A and B with ALPHA.
> >> +A and B should be lists (RED GREEN BLUE), where each element is
> >> +between 0.0 and 1.0, inclusive.  ALPHA controls the influence A
> >> +has on the result and should be between 0.0 and 1.0, inclusive."
> >
> > The doc string should document the return value.  Perhaps an example
> > of blending will also be a good addition to the doc string.
> 
> > From: Yuri Khan <yuri.v.khan@gmail.com>
> > Subject: Re: [PATCH] New function: color-blend
> > To: Joseph Turner <joseph@breatheoutbreathe.in>
> > Cc: Emacs Devel Mailing List <emacs-devel@gnu.org>
> > Date: Sat, 2 Nov 2024 18:26:57 +0700 (6 hours, 32 minutes, 20 seconds ago)
> > Flags: seen, personal
> > Maildir: /joseph-breatheoutbreathein/INBOX
> 
> > On Sat, 2 Nov 2024 at 14:03, Joseph Turner <joseph@breatheoutbreathe.in> 
> > wrote:
> 
> > > Here's `color-blend', a function to blend two colors into one.
> 
> > It would be nice to explicitly mention linear RGB space. Most
> > users[citation needed] these days are accustomed to colors specified
> > in sRGB space and naïve conversion dividing by 255 will not give the
> > desired blends. As an example, linear mixing black (0 0 0) with white
> > (1 1 1) yields (0.5 0.5 0.5). Naïve scaling will yield #808080 whereas
> > middle gray is somewhere around #bebebe in sRGB.
> 
> Thanks for the feedback!  Please see the attached revised patch.

Thanks, installed on master.

(I needed a few followup changes, as it looks like you ignored
byte-compilation warnings...)