On Sat, Jul 9, 2011 at 7:05 AM, Rob Beezer <goo...@beezer.cotse.net> wrote: > On Jul 7, 12:55 am, Francis Clarke <francis.w.cla...@gmail.com> wrote: >> All this should be simple to fix. > > Sounds good. With a fix for equality, there is still an outstanding > problem with comparisons of free modules. Presently, > > "Modules are ordered by their ambient spaces, then by dimension, then > in order by their echelon matrices." > > So again, there is an assumption that the echelon matrix is unique. > Is there a critical reason free modules should be ordered at all?
If there is no ordering on free modules, then any function that outputs a list of free modules is not deterministic -- the answer is randomly ordered. For testing purposes, this is a huge and annoying loss. Of course, it isn't at all critical for this that the ordering be the one used by __cmp__; one could define an ordering in some other function, which is used for ordering lists of submodules. We could instead have a method like M._useful_but_nanonical_total_ordering(N), then define __cmp__ like Nils suggests. -- William > > Rob > > -- > 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 > -- William Stein Professor of Mathematics University of Washington http://wstein.org -- 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
