On Wed, Jan 31, 2001 at 02:20:29AM -0200, Alexandre Oliva wrote:
> On Jan 30, 2001, "Markus F.X.J. Oberhumer" <[EMAIL PROTECTED]>
>wrote:
>
> > autoconf 2.49c (and the current CVS) fails if '.' is in PATH -
> > it tries to execute the directory `./m4' when searching m4.
I think the real problem here is having `.' further up your PATH than
`/usr/bin'. At worst autoconf could detect this and emit a warning
that everything might go horribly wrong unless `.' is removed from (or at
least moved to the end of) PATH.
> Yet another reason to not name the directory with m4 goodies `m4' :-(
>
> I vote for renaming it.
I disagree.
Specifically, sometimes a directory named m4 is all but required. For
instance in GNU m4, the sources for libm4 lie in a directory named m4,
since the header files contained therein are installed to $prefix/m4,
and use `#include <m4/error.h>' (for example), relying on
-I$top_srcdir to find the uninstalled error.h header in the m4
directory of the subtree, and allowing those same header files to work
from the installed tree without rewriting all of the #include lines.
In general, making certain directory names special for ease of
implementation is a bad idea. It took me a long time to figure out
why I couldn't name my @auxdir@ directory `aux' on cygwin, and I still
think that the reasons are stupid (DOS treats aux as a special device
file, so you can never name a directory aux under Windows).
Cheers,
Gary.
--
___ _ ___ __ _ mailto: [EMAIL PROTECTED]
/ __|__ _ _ ___ _| | / / | / /_ _ _ _ __ _| |_ __ _ ___ [EMAIL PROTECTED]
| (_ / _` | '_|// / |/ /| |/ / _` | || / _` | ' \/ _` | _ \
\___\__,_|_|\_, /|___(_)___/\__,_|\_,_\__, |_||_\__,_|//_/
home page: /___/ /___/ gpg public key:
http://www.oranda.demon.co.uk http://www.oranda.demon.co.uk/key.asc
