Op vrijdag 11 januari 2013 21:27:17 UTC+1 schreef Nils Bruin het volgende: > > > Isn't the real bug here that Matrix uses a much too heavy device to > keep track of a boolean flag (mutability=true/false)? Spinning this > off into a full-blown python class is just ludicrous for something as > fundamental as matrices.
I couldn't agree more. I assumed that "someone probably had a reason for that", but looking at the definition of the Mutability class (sage/structure/mutability.pyx), it's just ridiculous. (I guess it would be useful if multiple inheritance were possible in Cython, though). > If you rewrite that to be implemented via a > straight bint on matrix instances itself, the proper init code of > matrices will become much cheaper and perhaps not a hurdle for 2x2 to > call? In particular, it means the flag could be inited in the > __cinit__ of matrices and hence automatically be initialized. > > I'd assume there's very little code that depends explicitly on HOW > matrices keep track of their mutability. > > Compared to rewriting 2x2 to not inherit from Matrix (which I think > would indeed be necessary if you want to adhere to protocol: doesn't > inheriting mean relying on the initialization of the class you inherit > from?), making that change might not be so bad and it would benefit > much more of sage. > Agreed. Timo -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To post to this group, send email to sage-devel@googlegroups.com. To unsubscribe from this group, send email to sage-devel+unsubscr...@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel?hl=en.
