On Oct 19, 2006, at 11:19 PM, William Stein wrote: > Robert, > > We should make our matrices have nrows, ncols, and row iterations > and column iterations be of type size_t. On 64-bit systems, it > could "easily" happen that somebody wants to make a 2 x 2^33, > and we definitely want to support this. Plus, MAGMA doesn't > do this sort of thing well, so it's a chance for SAGE to shine:
That'd be one big matrix... but I agree. I'd also like to change the modn matrices to take advantage of 64-bit, is size_t the best way to get a machine word sized int? Also, is it signed? (I noticed at least one potentially "bad" subtractions in the modn code that currently used uint, and if we wanted to use balanced mod p values to reduce gather (eventually) we might want signed ints for that too.) >>> A := RMatrixSpace(Rationals(),2,2^30); > ^ > Runtime error in 'RMatrixSpace': Argument 3 (1073741824) is too large > >> A := RMatrixSpace(Rationals(),2,2^29); > > > > Magma: Internal error > Please mail this entire run [**WITH THE FOLLOWING LINES**] > to [EMAIL PROTECTED] > Version: 2.13-5 > You can print the whole input for this session by typing: > Magma: Fatal Error: Magma: Irrecoverable error; exiting > > > --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~----------~----~----~----~------~----~------~--~---