On Jan 5, 2010, at 6:38 AM, Sebastian Pancratz wrote:
For the purpose of making this a small and easy update, I think it
would be nice to not introduce new flags, provided one can get away
with it. I think the following policy might be OK:
- For exact rings with a GCD implementation, all user-callable
methods other than initialization (and perhaps suitable private helper
methods) should assume that the input is reduced and always return
reduced results. If the input is not reduced, the output should still
be mathematically correct, but might not be in reduced form.
- For exact rings without a GCD implementation, no assumptions are
placed on the input and no reduction takes place.
- For inexact rings, no assumptions are placed on the input and no
reduction takes place.
Does this seem reasonable?
Sounds reasonable to me.
Another ticket that touches a lot of the same code is http://trac.sagemath.org/sage_trac/ticket/7585
I'll try and rebase that on top of what you have in #7857.
As for avoiding reduction, I would either make it a property of the
Parent. Another (not necessarily exclusive) option would be to use
Python contexts, i.e.
F.<x> = Frac(QQ['x'])
with F.reduce(False):
print x/x
Printing x/x.
- 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
