dumpdef output stream

[Eric Blake]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

In 1.4.x, dumpdef used the trace mechanism for its output, so the -o
option or the debugfile macro could change the dump from stderr to some
other file (or discard it altogether).  In CVS head, revision 1.9 of
modules/m4.c (Aug 2001) changed to unconditionally use stderr
(http://lists.gnu.org/archive/html/m4-patches/2001-08/msg00016.html).
This was post m4-1.4o, but does not show up in NEWS.  Do we want to stick
with this change (meaning I fix my misdocumentation currently in head's
Invoking m4 node, and add a NEWS entry), or stay compatible with the
branch (meaning I revert the change)?  POSIX states that dumpdef goes to
stderr, but a POSIX-compliant usage of m4 will never encounter the -o
option or debugfile macro to change the debug stream to be different than
stderr, so that is not really an issue.

- --
Life is short - so eat dessert first!

Eric Blake             address@hidden
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.1 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFE7Eet84KuGfSFAYARAoCzAJ4mj31I8LnVwFTCfzASpnL5qcJXjQCdE3kU
/L7Tl8bBHAowGYo0Wgac1sA=
=WcvA
-----END PGP SIGNATURE-----