"David Abrahams" <dave@boost-consulting.com> wrote in message news:u3boli8k0.fsf@boost-consulting.com...
"Lucas Galfaso" <lgalfaso@gmail.com> writes:
Hi,
"Joel Eidsath" <jeidsath@gmail.com> wrote in message news:431931ED.7060901@gmail.com...
Do not agree. You should have ease of use _and_ efficiency. Modern C++ techniques allow for this.
I was trying to give myself wiggle room, but I pretty much agree with you. I don't expect anything as fast as GMP, but I imagine that a plain C++ library could still do very well.
Agree, I do not want to compete with GMP on performance, but, without sacrificing ease of use, have the best performance that a fully conformant C++ implementation allows.
IMO, if you don't want to compete with GMP, you should at least allow the library to be built as a wrapper over GMP.
Sorry if my statement was misunderstood, but when I said that I do not want to compete with GMP I did not mean that I am not considering speed, just that I will con write assembler to get a 5% improvement. The implementation I placed at the sandbox some few month ago, does have one of the fastest, pure C++, implementation of the modular exponentiation that I know about [and it has one of the poorest implementation of serialization/deserialization]. My goal in pursuing an implementation in Boost of big integers is not to have the _best_ implementation possible, but one that .) is really easy to use. .) fast (not fastest ever, just fast.) .) Boost like license. .) using C++ (not C as GMP) Regards, LG
-- Dave Abrahams Boost Consulting www.boost-consulting.com
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost