On Thu, May 15, 2008 at 1:05 PM, <[EMAIL PROTECTED]> wrote: > > David, your wording implies that there was one leak. Robert Miller and I sat > down for an hour a day every day for about 2 weeks, and fixed a number of > leaks every time. We eventually lost steam, because we didn't make an > appreciable dent in the number of memory leaks. Rather than 'leak', Leon > tends to 'splatter' memory. Picture a sopping wet sponge being thrown at a > wall. > > I haven't looked at A. Brouwer's patch, but after reading through much of > Leon's code, I have lost all faith that it can be trusted to produce a > correct result.
First, I appreciate your work on it! Second, i would greatly appreciate it if you could look over the Brouwer patch. My understanding of it is 0 but I was told that it makes Leon's code handle memory in a new way. Now, that could be good or bad:-) In any case, Leon's code is not used in this new fast min dist code, although they do share configure+make build scripts. > > > On Thu, 15 May 2008, David Joyner wrote: > >> >> Agreed. It is too slow. >> >> Guava now (very recently) has C code which computes min dist very fast >> for binary and >> ternary codes (codes from graphs are binary, so this would apply). However, >> it >> has not yet been linked to in SAGE. Why? There was a period when Michael said >> that the Leon code was leaking memory and he would not include the >> guava's compile+make in the sage build process. Now, that leak has (I >> think?) been fixed, >> thanks to a patch of A. Brouwer. ANow, the new C code for the fast min >> dist is written >> by Cen Tjhai, not Leon, but it uses the same compile+make scripts. >> >> I have just checked and as of sage 3.0, this binary is included in SAGE. So, >> to >> do a fast min dist computation *in GUAVA*, use the MinimumWeight >> command >> http://sage.math.washington.edu/home/wdj/guava/htm/chap4.html#X84EDF67B86B4154C >> Obviously, I should rewrite sage's C.minimum_distance function to call >> this faster one >> very soon. >> >> >> >> On Thu, May 15, 2008 at 9:35 AM, William Stein <[EMAIL PROTECTED]> wrote: >>> >>> Hi David, >>> >>> Here is some "non-fan mail" about codes in Sage from Africa. Any >>> comments? >>> >>> ---------- Forwarded message ---------- >>> From: Khumbo Kumwenda <[EMAIL PROTECTED]> >>> Date: Fri, May 9, 2008 at 12:24 PM >>> Subject: used sage >>> To: [EMAIL PROTECTED] >>> >>> >>> Hello, >>> >>> I have used sage in an essay I am writing on linear codes from >>> triangular graphs and their complements. It was OK but, as other >>> programming languages I have used, calculating the minimum weights of >>> codewords takes too long when you have a lot of codewords. Not sure if >>> there is a way out. >>> >>> Khumbo Kumwenda >>> African Institute for Mathematical Sciences >>> Muizenberg >>> Cape Town >>> South Africa >>> >>> >>> >>> -- >>> William Stein >>> Associate Professor of Mathematics >>> University of Washington >>> http://wstein.org >>> >>>> >>> >> >> > >> > > > > > > --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
