> < ^ Date: Tue, 12 Oct 1999 17:01:31 +0100
^ From: Joe Fields <fields@math.uic.edu >
> < ^ Subject: Re: guava question

Hi,

I don't think it's a memory issue.

I've run in to several similar errors with
the guava functions that interface with
Jeff Leon's library of C programs (ConstantWeightSubcode
is one such). The problem seems to be in
the communication process between the main
guava functions and the C programs which
is accomplished via the file system. That is
the C program is induced to place it's output
in a file whose contents look like:

GUAVA_TEMP_VAR := (whatever the output of Leon's utility is);

Leon's programs expect file name arguments that are
relatively short (I forget, maybe 8 or 9 characters long...)
The names that Guava cooks up for these temp files are
often longer (it's system dependent). I'd try hard-coding the temporary
file names in your Guava libraries (but use a fairly esoteric name so you
don't overwrite something you actually need!)

I think what's going on is that Leon's program is not
getting reasonable input and so the contents of that file
which communicates back to GAP is:

GUAVA_TEMP_VAR := FALSE;

So you end up trying to multiply a finite field element "i"
with a boolean "GUAVA_TEMP_VAR" in one of the lines below.
(The error messages aren't always informative, but I think
the one below is right.)

Best,
Joe Fields

At 09:51 PM 10/12/99 -0400, you wrote:
>Hello, I'm running into what looks like memory problems with the
>'ConstantWeightSubcode' function in Guava (I'm running gap 3.4.4 in
>DOS). Or maybe I've missed something.
>
>Here's an example:
>
>gap> lc5 := LexiCode(5,3,GF(2));
>a linear [5,2,3]2 lexicode over GF(2)
>gap> Elements(lc5);
>[ [ 0 0 0 0 0 ], [ 0 0 1 1 1 ], [ 1 1 0 0 1 ], [ 1 1 1 1 0 ] ]
>gap> cwsc := ConstantWeightSubcode(lc5,3);
>Error, operations: product of finite field element and boolean is not
>defined \
>at
>Append( Els, i * GUAVA_TEMP_VAR ) ... in
>arg[1].operations.ConstantWeightSubcode( arg[1], arg[2] ) called from
>ConstantWeightSubcode( lc5, 3 ) called from
>main loop
>
>Any help appreciated -- offlist replies ok.
>
>Drew Krause
>
>
Joe Fields
Mathematics Department
Southern Connecticut State University
New Haven CT  06515
fields@scsu.ctstateu.edu
(203) 392-6519
Fax: (203) 392-6808

> < [top]