On Wednesday, August 12, 2015 at 2:49:26 PM UTC-7, Volker Braun wrote: In many cases where we can solve the word problem we just use canonical > representatives from the get-go. So there == would still do what you would > naively expect even when comparing presentations in a given parent. But > when we can't we should't pretend that we can magically solve the word > problem. I'd take something that lets me write effective algorithms any > time over an ideologically pure but useless implementation. >
But we don't in the example that Nathann points out! In the finite FP group < a | a^4 > of course we can solve the word problem, and Gap uses it to answer equality questions (via an isomorphism to a permutation group, if I understand their documentation). Yet we do NOT rewrite group elements into canonical form. Do you really just want equality and hash on FP groups to be what we have on the covering free presentation? GAP doesn't do that itself (and as a result computations there can lead to unbounded running times and outright failures). I would propose that we follow the library that we wrap and do the same. For hashing that means that we must get a hold of the canonical form that GAP computes internally somewhere, though. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
