ECL update



James Amundson wrote:
> I just:

Hi James,

> 1) Fixed the shadowing problem I mentioned in the previous message.
> 
> 2) Fixed a problem with the maxima script getting overwritten by ECL.
> 
> 3) Fixed command-line arguments for ECL.
> 
> I think ecl-maxima is now fully functional. The primary remaining issue 
> is that I do not think incremental builds work. I don't know how to fix 
> that.

I just checked out Maxima CVS and build it with ecl-0.9l on x86-64. I am 
seeing two test failures:

Running tests in rtest2:
********************** Problem 40 ***************
Input:
                      3                3
determinant(hessian(y  - 3 a x y z + x , [x, y, z]))


Result:
Load failed for 
/home/mabshoff/maxima+ecl/maxima/share/linearalgebra/eigens-by-jacobi.lisp
Load failed for 
/home/mabshoff/maxima+ecl/maxima/share/linearalgebra/load-linearalgebra-lisp-files.lisp
error-catch

This differed from the expected result:
       2  3       3                      2      2
- 54 a  x  - 27 a  x y z - 3 a y (18 a y  + 9 a  x z)

********************** Problem 41 ***************
Input:
                               2
subst(1, z, quotient(%, - 54 a ))


Result:
0

This differed from the expected result:
  3            3
x  + a x y + y

54/56 tests passed.


> --Jim

I used an external boehm_gc (7.1) and did not the build in one provided 
by ecl (6.8). There is a bug in the ecl build system when using an 
external boehm_gc that is not installed in the the default location, but 
I will report that issue to the ecl mailing list. Running the ecl ansi 
test suite resulted in the couple expected failures:

32 out of 21696 total tests failed: COMMON-LISP-PACKAGE-NICKNAMES, 
COMMON-LISP-USER-PACKAGE-NICKNAMES,
    BOUNDP.5, PROCLAIM.ERROR.7, CTYPECASE.15, CTYPECASE.16, EQUALP.5,
    EQUALP.6, DEFGENERIC.ERROR.20, DEFGENERIC.ERROR.21,
    CALL-NEXT-METHOD.ERROR.1, CALL-NEXT-METHOD.ERROR.2,
    DEFMETHOD.ERROR.14, DEFMETHOD.ERROR.15, MAKE-CONDITION.4,
    UPGRADED-ARRAY-ELEMENT-TYPE.8, BIGNUM.FLOAT.COMPARE.7,
    BIGNUM.FLOAT.COMPARE.8, EXP.ERROR.10, EXP.ERROR.11, MAP.ERROR.11,
    ALL-STANDARD-GENERIC-FUNCTIONS-ARE-INSTANCES-OF-THAT-CLASS,
    FORMAT.LOGICAL-BLOCK.23, SYNTAX.DOT-ERROR.1, COMPILE-FILE.2,
    COMPILE-FILE.2A, COMPILE-FILE.6, COMPILE-FILE.6A, COMPILE-FILE.8,
    COMPILE-FILE.14, COMPILE-FILE.18, MISC.521.


You did mention that Maxima on top of ecl failed for you on some Fedora 
box on x86-64 due to out of memory conditions. I do remember similar odd 
problems with python on FC 8 x86-64 where reverting to a previous kernel 
fixed the issue. This might or might not be the problem you are having, 
so consider this a shot in the dark :)

Thanks for fixing the Maxima+ECL issues. Hopefully I can contribute a 64 
bit MSVC installer for Maxima 5.16+ECL once the final release has been done.

Cheers,

Michael

> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima
>