Ben Pfaff <blp@xxxxxxxxxxxxxxx> writes: > Russ Allbery <rra@xxxxxxxxxxxx> writes: >> Practically speaking, I suspect that you'll improve portability to >> systems that people actually care about by not worrying about >> strings.h at all and not even trying to include it. I think we've >> now reached the point where it's more likely that an OS will provide >> a broken strings.h out of a misguided sense of backwards >> compatibility than that someone will really want to build new >> software on SunOS. > I think that <strings.h> is still the only header in which POSIX > declares ffs(), strcasecmp(), and strncasecmp(). That's a reasonable > reason to #include <strings.h>, although Konstantin may not need those > functions anyhow. Huh. I've never seen a system that didn't declare strcasecmp() and strncasecmp() in string.h. But indeed, that's what POSIX documents. Thanks! I didn't realize that. -- Russ Allbery (rra@xxxxxxxxxxxx) <http://www.eyrie.org/~eagle/> _______________________________________________ Autoconf mailing list Autoconf@xxxxxxx http://lists.gnu.org/mailman/listinfo/autoconf
