[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Saturday 22 April 2006 8:49 am, Theo Zourzouvillys wrote: > On Friday 21 April 2006 22:05, Neil Williams wrote: > > Would someone with a 64-bit system mind testing QOF 0.6.4 from > > SourceForge with the attached patch and telling me if it compiles OK? (It > > fails without the patch). Other systems are fine, it's only 64bit systems > > that are relevant at this time. > > No it does not compile with the patch. Still improving the patch. I've sent another off-list to you, Robin and David. Thanks all for your assistance. Looks like my next purchase will be a 64bit system! The latest patch is attached for anyone else testing. ;-) Any recommendations for a ready built 64bit system? Naturally, Debian will be installed upon purchase if not before. > if you are refering to types as "long long int", you will want a format > string of %lld (or %qd on osx iirc) - alternatively take a look at > <inttypes.h>. I'm using glib macros wherever possible - they implement inttypes.h in a familiar manner. As Simon hinted, this is primarily for compatibility, not just with GnuCash but to provide a consistent "feel" across QOF applications. QOF is gradually moving all definitions to their glib equivalents, including gint and gchar. No point re-inventing the wheel, use the portability support already available in the main library that QOF uses itself. > also, your configure.in is dropping my personal CFLAGS i provide when > calling ./configure Hmm. I'll check that for the next upstream release. Can't fix that right now. Did you check the Makefile CFLAGS? It may just be that my AC_OUTPUT is misleading (it outputs $warnflags which is derived from CFLAGS when it should also output the original CFLAGS). -- Neil Williams ============= http://www.data-freedom.org/ http://www.nosoftwarepatents.com/ http://www.linux.codehelp.co.uk/
Attachment:
test-date.c_64-4.patch
Description: Text Data
Attachment:
pgp9yVSHmp8O7.pgp
Description: PGP signature