Re: Directory structure changes in Emacs installed on MS-Windows

[Eli Zaretskii <eliz@gnu.org>]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Cc: Stephen Leake <stephen_leake@member.fsf.org>,  emacs-devel@gnu.org
> Date: Fri, 05 Apr 2013 09:19:21 -0400
> 
> > It would be important if it were true, but I don't think it's true.
> > The way epaths.h is generated from epaths.in on Posix platforms leaves
> > the root of the Emacs installation tree (specified via --prefix)
> > hard-coded into the binary, and Emacs (AFAIK) currently doesn't
> > support relocation of the installation tree on Posix systems.
> > By contrast, on Windows relocation is a matter of routine.  So I cannot
> > use the epaths-force target of the top-level Makefile anyway.
> 
> FWIW, such a relocation issue came up with the NS port, and as
> I pointed out back then, it would be better to do it once and for all
> uniformly, i.e. also do that relocation for posix platforms.

I believe you are talking about this:

  http://debbugs.gnu.org/cgi/bugreport.cgi?bug=11959#53

No one picked up that gauntlet.  One piece of the puzzle that is still
missing (at least as far as I'm concerned) for its solution is how to
reliably determine on Posix platforms the directory where the invoked
Emacs executable lives.  I vaguely remember that the value of argv[0]
is unreliable, and therefore even searching for it on PATH won't do.
How can that be done reliably?