On 1 Apr., 17:58, Fredrik Johansson <fredrik.johans...@gmail.com>
wrote:
> I'm proud to announce the initial release of FLINT 3.0. The FLINT 3
> series is a complete rewrite of FLINT (Fast Library for Number Theory)
> from scratch, like FLINT 2.x was a complete rewrite of FLINT 1.x
> before it.
>
> FLINT 3.0 is available as an interactive interpreter 
> at:http://www.flintlib.org/flint3/
>
> The most significant change in FLINT 3 is that the implementation
> language has been changed from C to JavaScript. With modern
> JIT-compiling JavaScript engines, antiquated static languages such as
> C are no longer necessary for high performance computing. Switching to
> JavaScript allows us to make the transition to emerging computing
> platforms such as Windows 8, tablets, mobile phones, and the cloud.
>
> A second major change is that the only supported coefficient ring is
> double-precision floating-point numbers (with JavaScript semantics).
> Exact bigints and rationals are just too slow and complicated to be
> worth the trouble, and besides, you rarely need more than 16 digits of
> precision anyway. Most of the scientific computing world has already
> realised as much, and we anticipate that this bold move will give us a
> head start of 6-8 years over the rest of the computer algebra
> community.
>
> The FLINT 3.0 interpreter uses a Lisp-like syntax. As an example
> computation, type 25 or (1 + 1) followed by ENTER.

There's still some work to be done to get it properly integrated into
William's Lisp port of Sage.

But so far looks very promising.


-leif

> Since this is a
> first release, some minor bugs are still present, and the code might
> not work in all web browsers. We hope to sort these issues out in the
> coming years using iterative development.
>
> --Fredrik, on behalf of the FLINT development team

-- 
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

Reply via email to