Thanks. This fixes the bug I described.
But now, I have another problem, during "make":
Compiling deliverquota.c
gcc -traditional-cpp -Wall -I./.. -I.. -o deliverquota deliverquota.o libmaildir.a ../rfc822/librfc822.a ../numlib/libnumlib.a
ld: ../numlib/libnumlib.a(strhdevt.o) literal C string section (__TEXT,__cstring) does not end with a '\0'
ld: ../numlib/libnumlib.a(strhinot.o) literal C string section (__TEXT,__cstring) does not end with a '\0'
make[2]: *** [deliverquota] Error 1
make[1]: *** [all] Error 2
make: *** [all-recursive] Error 1
I add a "\0" to the next constant, but the same error arises:
static const char rcsid[]="$Id: deliverquota.c,v 1.21 2003/01/20 13:47:19 mrsam Exp $";
->
static const char rcsid[]="$Id: deliverquota.c,v 1.21 2003/01/20 13:47:19 mrsam Exp $\0";
Another question: are the global variables (RANLIB ... CPPFLAGS) of general use when you compile *any* package to OSX? Or are they specific to this build?
Sorry, but I must yet use the --enable-mimetypes; otherwise, ./configure reclaims it... (my build is "sqwebmail-3.5.0.20030301.tar.gz").
Actually, I'm not sure whether the last one is needed for sqwebmail, or if it's specific to some other part of Courier. (Personally, I always build the full system, not just sqwebmail... and I forget which problem it solved. =) But the first 3 are *definitely* needed if you want to have any chance of successfully compiling on OS X.
./configure \ 'RANLIB=ranlib -c' \ 'CFLAGS=-traditional-cpp' \ 'CXXFLAGS=-traditional-cpp' \ 'CPPFLAGS=-D _INTL_REDIRECT_MACROS'
I would also point out that you are using an improper specification of the imagedir path. That path should be based on the web server's document root, not the filesystem root. E.g. '--enable-imagedir=/images/sqwebmail'
And lastly... There shouldn't be any need for the --enable-mimetypes setting anymore on OS X. The Mac OS X default location for mime.types was added to the Courier build system in a previous release. So configure should find it automatically. -- Of course, if it doesn't, then ask Sam if there has possibly been a regression of some sort with regard to that option.
Best regards, Guy
