1 |
Bump request was open for a while already: |
2 |
https://bugs.gentoo.org/show_bug.cgi?id=542682 |
3 |
I bumped it yesterday, is the ntl-62 compat patch essential? |
4 |
Patrick reported a test failure that I cannot reproduce yet: |
5 |
https://bugs.gentoo.org/show_bug.cgi?id=545378 |
6 |
Do you know anything about it? |
7 |
|
8 |
On 03/04/15 00:53, François Bissey wrote: |
9 |
> Speaking of flint would you mind importing 2.4.5 from the sage-on-gentoo |
10 |
> overlay to the main tree. I should have opened a bump request last month |
11 |
> when it was released. |
12 |
> |
13 |
> François |
14 |
> |
15 |
>> On 3/04/2015, at 01:33, Thomas Kahle <tomka@g.o> wrote: |
16 |
>> |
17 |
>> Bumped, but now flint became incompatible: |
18 |
>> |
19 |
>> https://github.com/wbhart/flint2/issues/131 |
20 |
>> |
21 |
>> On 28/03/15 20:54, François Bissey wrote: |
22 |
>>> And Victor just announce ntl 9.0 on sage-devel: |
23 |
>>> |
24 |
>>> With much trepidation, I have introduced a (hopefully minor) |
25 |
>>> backward incompatibility into NTL. |
26 |
>>> |
27 |
>>> The interface to the single-precision modular arithmetic |
28 |
>>> routines has been modified slightly. |
29 |
>>> This interface change allows for more flexible and more |
30 |
>>> efficient implementation of these routines, |
31 |
>>> which play a crucial role at many levels in NTL. |
32 |
>>> |
33 |
>>> Basically, these changes to the interface abstract away |
34 |
>>> some implementation details that arguably should never been there |
35 |
>>> in the first place. |
36 |
>>> By coding to the new interface, NTL clients will be able to |
37 |
>>> benefit from the current and future improvements. |
38 |
>>> |
39 |
>>> In particular, on 64-bit x86/GCC platforms, single precision |
40 |
>>> moduli can now be up to 60 bits, rather than 50 bits. |
41 |
>>> While some operations may in fact be a little slower, the most important |
42 |
>>> ones (like MulModPrecon) should not be. |
43 |
>>> Using larger moduli speeds up a number of things, like ZZ_pX |
44 |
>>> arithmetic, as fewer primes need to be used in Chinese Remaindering steps. |
45 |
>>> Other applications benefit from larger moduli as well. |
46 |
>>> |
47 |
>>> It is expected that most NTL clients will not be affected at all. |
48 |
>>> Moreover, any code that needs to be updated will be detected |
49 |
>>> by the compiler, and the updates should be simple and mechanical. |
50 |
>>> There is also a configuration flag that will enable the legacy |
51 |
>>> interface (although this is not recommended practice). |
52 |
>>> |
53 |
>>> For more, go to http://www.shoup.net/ntl |
54 |
>>> |
55 |
>>> |
56 |
>> |
57 |
>> -- |
58 |
>> Thomas Kahle |
59 |
>> http://dev.gentoo.org/~tomka/ |
60 |
>> |
61 |
> |
62 |
> |
63 |
|
64 |
-- |
65 |
Thomas Kahle |
66 |
http://dev.gentoo.org/~tomka/ |