it still fails. I think Joel's analysis is correct, the configure test seems to be wrong. What is this bit of code in the configure supposed to catch? { /* SCO 3.2v4 cc rejects this. */ char *t; char const *s = 0 ? (char *) 0 : (char const *) 0; *t++ = 0; } randolph
again I don't think so. We could remove this test but: it is *not* executed during the configure test. The mere compilation must never segfault. I don't exactly know what it tests. But there must be some reasons for this test to be in the configure of all mozilla applications. I guess they use this pattern somewhere in code.
The test is the *same* as its for mozilla-firefox, mozilla & mozilla-thunderbird. Again, we have successful builds by joel and carlos. What shall we do??
Alex -- PGP messages preferred. | .''`. ** Debian GNU/Linux ** Alexander Sack | : :' : The universal asac@jwsdot.com | `. `' Operating System http://www.jwsdot.com/ | `- http://www.debian.org/