It turns out that there is an Itanium-only (?) bug in GAP's garbade- collection that breaks SaveWorkspace(), that can be reproduced in GAP 4.4.12. (4.4.10 seems to have identical code in that part, but the bug does not seem to manifest itself there for some reason) so we have to wait for GAP kernel people to fix this, before this can go ahead.
Dmitrii On Jan 10, 4:31 am, Minh Nguyen <nguyenmi...@gmail.com> wrote: > Hi Dmitrii, > > On Sun, Jan 10, 2010 at 3:25 AM, Dima Pasechnik <dimp...@gmail.com> wrote: > > <SNIP> > > > OK, so let me repost here the state of affairs AFAIK. > > For some reason it is believed that GAP.4.4.12 cannot be currently > > packaged with Sage, due to a problem with an Itanium build. > > Sage is known to successfully build on the following Itanium platforms with > GCC: > > * Red Hat Enterprise Linux Server 5.3, IA-64 Itanium 2 @ 1.5GHz, GCC 4.4.2 > > * SUSE Linux Enterprise Server 10 SP1, IA-64, GCC 4.4.2 > > A list of all known platform/hardware combinations on which Sage 4.3 > is known to successfully build is found on the Sage wiki [1]. > > > Is the Intel 10.1 compiler a must? > > No. Currently, the Intel C/C++/Fortran compilers are not required to > build Sage. Other than GCC, I don't know of any other compiler that > has been used to successfully build Sage. > > > If not, what does stop GAP 4.4.12 > > then? > > GAP is a standard package which is distributed with every version of > Sage. For any update to the GAP package, one needs to ensure that Sage > together with the updated GAP package successfully compile on the > compile farm. You could make the updated GAP package available online > somewhere and ask people to test it out. > > [1]http://wiki.sagemath.org/devel/BuildFarm/sage-4.3 > > -- > Regards > Minh Van Nguyen
-- 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
