Re: a review of the merge (Re: Emacs.app merged)

[Jason Rumney <jasonr@gnu.org>]

Adrian Robert wrote:

> Hi Jason,
> 
> Do you have any objection to moving/renaming w32-load-color-file to
> xfaces.c so the NS port can use it?  Also, does it need to be exposed to
> lisp (syms_of...)?  It's not called from anywhere in the emacs tree lisp
> anyway.

No objection to renaming, I guess the export to lisp is so users can
override the color names with their own rgb.txt file, though I suspect
the number of users who do so is pretty close to zero.

> Finally, how necessary is the hard-coded set of xcolors definitions in
> w32fns.c?  It's not OK just to abort if this file is not found in
> data-directory?

I think Emacs should not abort (on any platform) if it does not find
data files. Emacs 22 starts up on Windows if given -nw without ANY other
files, without -nw it requires lisp/term/w32-win.el, but in the trunk
that is dumped, so theoretically it should be possible to do basic
editing with just an executable (currently it is producing an "Invalid
code(s)" error).