On 9 February 2012 09:38, Jeroen Demeyer <jdeme...@cage.ugent.be> wrote:
> On 2012-02-09 10:34, David Kirkby wrote: > > No, one what would do in this case is > > > > ./configure CC="gcc -m64" > > Exactly. If every spkg would actually respect the CC environment > variable, this would work and we wouldn't need SAGE64 anymore. > I would not be 100% sure about that. I think there would be the odd exception. bzip2 I believe needs a special flag for a 64-bit build (something like --64). But the odd exceptions should be easy to resolve. There are a lot of packages in Sage which don't respect CC. I've fixed some of them, but gave up at some point as there were too many. Is is funny, as there are the odd package which has in it. CC=cc then some Sage developer had changed it to CC=gcc Then I've come along and changed the code so it actually respects CC. It's the quality of some of the C code in particular which makes me worry a bit about Sage (see my previous comment on this thread). It's clear there is C code in Sage, where the developer does not have an ****ing clue about how to write C. In one case, another Sage developer said the code was "virtually obfuscated", but I found it would not be eligible for the "obfuscated C contest" http://www.ioccc.org/ as the code is not even legal C. It just happens gcc will compile it. Dave -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
