Re: Compiling emacs-27.0.91 on macOS Catalina

[Alan Third <alan@idiocy.org>]

On Fri, May 15, 2020 at 12:56:25PM +0300, Eli Zaretskii wrote:
> > Date: Fri, 15 May 2020 07:25:58 +0100
> > From: Alan Third <alan@idiocy.org>
> > Cc: paranda@it.uc3m.es, emacs-devel@gnu.org
> > 
> > Alternatively we could check the results from trying to load the file.
> > There's no way I can see to catch the error, but we can count the
> > number of colours available. Is this too late for Emacs 27?
> 
> I don't think I understand the effect of this change yet, so I cannot
> answer the question.  How will this prevent the problem?

The color file is simply a native version of etc/rgb.txt. It contains
colour values and their names.

I'm not really sure why we save it, to be honest.

On first run Emacs detects there is no color file (cl == nil in the
code) and generates it from rgb.txt.

Since we're unable to detect directly if loading the color file has
thrown an error my change checks whether there are less colors loaded
than there are in rgb.txt. If there are less then it will regenerate
the file.

It's not great, rgb.txt could change, but I assume at this stage that
that's unlikely or rare.
-- 
Alan Third