On Tue, Dec 30, 2008 at 12:02 PM, Duane Ellis <[email protected]> wrote: > Ted> The test in configure.in is testing for linkage against a function > Ted> that openocd is not even using. In my opion, that makes the test is > Ted> broken. > > Be that as it may - and I generally agree with you - Version number hell > sucks. None of us have means (and time, energy, etc) to go backwards and > test permutations from Version Number Hell. > > But - as it stands now - the Win32 version I believe - requires at least > 0.12 etc.. See the "CONTRIB" folder for info about this. So what are we to > do? > > This entire problem is - interesting - and painful. as I pointed out in my > original mail that Steve Franks quoted.. > > duane >> I wonder how many other bumps we will find :-( > > Our goal (putting blinders on and being 100% totally openocd centric) - is > to get to a release 1.0 milestone. > > Having FreeBSD + MAC + WINDOWS + LINUX = all supported @ 1.0 is a great > thing.
It appears that FreeBSD uses 0.10 also. http://www.freebsd.org/cgi/ports.cgi?query=libftdi&stype=all Thus, no amount of tweaking of CFLAGS and/or configure options will allow that test to pass. My patch should allow Steve to get openocd to get past the configure stage on FreeBSD (assuming that /usr/local/include is in his header search path and /usr/local/lib in the lib search path). Ted Roth _______________________________________________ Openocd-development mailing list [email protected] https://lists.berlios.de/mailman/listinfo/openocd-development
