This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: ok to use autogen to generate Makefile.in?
- From: Stan Shebs <shebs at apple dot com>
- To: Nathanael Nerode <neroden at doctormoo dot dyndns dot org>
- Cc: gdb at gdb dot gnu dot org, gcc at gcc dot gnu dot org
- Date: Fri, 05 Apr 2002 12:46:23 -0800
- Subject: Re: ok to use autogen to generate Makefile.in?
- References: <20020405061232.GA24166@doctormoo.dyndns.org>
Nathanael Nerode wrote:
>
> So what I want to know is, is it acceptable to use autogen to create
> Makefile.in? It generates a nice, human-readable Makefile.in from a
> nice, human-readable-and-maintainable template. But it does use
> autogen, which is not a traditional Makefile generation tool.
autogen needs some portability work; it doesn't build on Mac OS X
for instance. After some dinking around trying to get it to go,
just so I could do a fixincludes thing, I ended up faking the
header to be fixed on a Linux system and doing the work there.
It also has a guile dependency, so you have to have guile installed
on the system. My RH 7.2 machine did not have the config script
preinstalled even though it had the guile executable, so I had to
download and install guile before I could build autogen.
The top-level Makefile.in doesn't have to be regenerated very
often, but I'd like to see autogen in better shape before
making myself dependent on it for anything.
Stan