On Feb 16, 2010, at 12:18 AM, François Bissey wrote:
On Tue, 16 Feb 2010 12:41:31 William Stein wrote:
If you link against mpir then the above discrepancy would go away.
However, i can also tell you that the above discrepancy is harmless.
The issue is simply that both answers are right. e.g., the line
sage: x = crt(2, 1, 3, 5); x
uses the Chinese Remainder Theorem to find an integer x such that x %
3 == 2 and x%5 == 1. Both -4 and 11 are correct choices of x.
The other related tests involving "modular symbols" are different
because of the same function.
I guess a finer point to know is. is this result correct but not
reduce
to same domain? And does it have nasty consequences later - ie do
other thing rely on the result to be in this domain?
No, there is definitely no nasty consequence.
One more question, if we need to link against mpir's libgmp, does
linking
sage itself against it enough, or do we need to also link pari,
mpfr,
mpfi and other against it as well? mpfr in itself would be a
problem as
it is part of the toolchain (gcc is compiled against it).
I think linking sage itself is likely to be enough for doctests to
pass, though it is hard to be sure. It would be nice to link pari
too.
I just had a thought. why is sage still linked to libgmp rather than
libmpir?
There is nothing in libgmp that isn't in libmpir - or am I mistaken?
MPIR is supposed to be a drop in replacement for GMP and provides a
"compatibility" mode, so we really are linking to MPIR. (It's also
force of habit, and easier than patching all the upstream packages).
- Robert
--
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
