This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: top level: make more dependencies explicit
- From: Zack Weinberg <zack at codesourcery dot com>
- To: Andrew Cagney <ac131313 at redhat dot com>
- Cc: Nathanael Nerode <neroden at doctormoo dot dyndns dot org>, gdb at sources dot redhat dot com,binutils at sources dot redhat dot com, gcc at gcc dot gnu dot org
- Date: Sun, 29 Sep 2002 12:25:04 -0700
- Subject: Re: top level: make more dependencies explicit
- References: <20020929165232.GA27545@doctormoo.dyndns.org> <3D9733C2.2010405@redhat.com> <20020929172608.GA27678@doctormoo.dyndns.org> <3D973C44.6090601@redhat.com> <20020929174544.GA30373@doctormoo.dyndns.org> <3D974828.4050009@redhat.com>
On Sun, Sep 29, 2002 at 02:36:24PM -0400, Andrew Cagney wrote:
> What does the GNU coding standard have to say about the release process?
There are some Makefile targets defined for this purpose, but
personally I think that part of the spec is out of date and should be
ignored.
> I'd also be wary of a ``rewrite'', the top-level stuff iteracts with
> sub-directories in strange ways. I think reserving the existing
> behavior (but perhaphs outside of the Makefile.in) would be a better
> incremental step.
>
> Also, how does this compare to GCC's release process.
GCC uses a number of scripts (such as maintainer-scripts/gcc_release)
kept in CVS, but separate from the Makefiles. I think this is the way
to go. It prevents exactly this sort of interaction problem. Also,
the version control policy on release scripts is likely to be
different enough that isolation in separate files will make life
easier. And it's ever so much easier to write big shell scripts if
you don't have to put backslashes at the end of every line.
zw