1 |
Bumped, but now flint became incompatible: |
2 |
|
3 |
https://github.com/wbhart/flint2/issues/131 |
4 |
|
5 |
On 28/03/15 20:54, François Bissey wrote: |
6 |
> And Victor just announce ntl 9.0 on sage-devel: |
7 |
> |
8 |
> With much trepidation, I have introduced a (hopefully minor) |
9 |
> backward incompatibility into NTL. |
10 |
> |
11 |
> The interface to the single-precision modular arithmetic |
12 |
> routines has been modified slightly. |
13 |
> This interface change allows for more flexible and more |
14 |
> efficient implementation of these routines, |
15 |
> which play a crucial role at many levels in NTL. |
16 |
> |
17 |
> Basically, these changes to the interface abstract away |
18 |
> some implementation details that arguably should never been there |
19 |
> in the first place. |
20 |
> By coding to the new interface, NTL clients will be able to |
21 |
> benefit from the current and future improvements. |
22 |
> |
23 |
> In particular, on 64-bit x86/GCC platforms, single precision |
24 |
> moduli can now be up to 60 bits, rather than 50 bits. |
25 |
> While some operations may in fact be a little slower, the most important |
26 |
> ones (like MulModPrecon) should not be. |
27 |
> Using larger moduli speeds up a number of things, like ZZ_pX |
28 |
> arithmetic, as fewer primes need to be used in Chinese Remaindering steps. |
29 |
> Other applications benefit from larger moduli as well. |
30 |
> |
31 |
> It is expected that most NTL clients will not be affected at all. |
32 |
> Moreover, any code that needs to be updated will be detected |
33 |
> by the compiler, and the updates should be simple and mechanical. |
34 |
> There is also a configuration flag that will enable the legacy |
35 |
> interface (although this is not recommended practice). |
36 |
> |
37 |
> For more, go to http://www.shoup.net/ntl |
38 |
> |
39 |
> |
40 |
|
41 |
-- |
42 |
Thomas Kahle |
43 |
http://dev.gentoo.org/~tomka/ |