FPLLL Days 5

Skip to first unread message

Martin R. Albrecht

Oct 14, 2019, 7:56:28 AM10/14/19
to Fplll-Devel
Hi all,

FPLLL Days 5 https://github.com/fplll/fplll/wiki/fplll-days-5 are upon us in two weeks.

1/ I suggest we meet at 10am on Monday in CDT Room in the Bedford Building for introductions and project distribution (i.e. who plans to work on what)

2/ Speaking of which, there are a few project ideas on the Wiki https://github.com/fplll/fplll/wiki/fplll-days-5 but given the large expected turnout, it might be a good idea to have a few more ideas. Thus, feel free to suggest ideas here or on the Wiki page.

3/ Also, we did a few tutorial sessions at FPLLL Days 3 https://github.com/fplll/fplll/wiki/fplll-days-3. Please let us know if something like this (and what) would be useful to you at FPLLL Days 5. We don’t want to fill the programme with talks, but a few tutorials might be possible.



_pgp: https://keybase.io/martinralbrecht
_www: https://malb.io

Christian Peel

Oct 20, 2019, 11:52:13 PM10/20/19
to Martin R. Albrecht, Fplll-Devel

I'd be happy for tutorials; I'd enjoy a tutorial on G6K and other more basic topics required to understand it (Gauss sieve, other variants of sieving) or other basic tutorials.

I mentioned to you that I was interested in making a Julia-language wrapper around fplll. After looking at it, I now think that making a fully-featured wrapper is a bigger project than I am looking for.    I managed to call fplll in a not-so-clean way from Julia, and make a few comparisons between fplll, the LLL in Nemo (Flint) [1], and the Julia LLL  in LLLplus [2].  The code and comparisons are available at [3], including a few comments on what would be required to implement a good Julia wrapper around fplll.  

My interest in lattice tools is to educate myself; adding features to LLLplus seems sufficiently educational for now, even though it is much slower and will never be as fully-featured as fplll.  During fplll-days I plan to work on adding an L2-style LLL and a BKZ to LLLplus. I'm also interested in a simple sieve such as [4], though that may be too ambitious.


You received this message because you are subscribed to the Google Groups "fplll-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fplll-devel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fplll-devel/87blujczh2.fsf%40road.
Reply all
Reply to author
0 new messages