Work at SourceForge, help us to make it a better place! We have an immediate need for a Support Technician in our San Francisco or Denver office.
Close
From: Nikolas Kauer <kauer@ph...>  20000829 18:22:12

Hi, this post refers to option 2 of RE: [clisplist] punimax: > 2. if you would rather use the real thing (maxima), you should > > A. visit http://www.ma.utexas.edu/users/wfs/maxima.html > and get the current CVS version of maxima > > B. apply the appended patch (BEWARE: work in progress!) > > C. get the latest version of CLISP (http://clisp.cons.org) > > D. follow the instructions in maxima/README I installed clisp20000306 with the clispalphadecosf4.0d.diff patch on our Alpha cluster running Tru64 Unix 4.0E, and the test suite ran fine. I then followed the above instructions and tried to install maxima for one of our postdocs. The patch worked fine on a recent CVS version (except for src/homog.lisp, where the indentation changed slighly). The makefile worked fine, except that the file src/dumpclisp.lisp was missing. When compiling maxima I got lots of warnings, but no errors. I used:  (load "sysdef.lisp") (make:make :maxima :compile t) (lisp:saveinitmem "maxima.mem" :quiet t :initfunction 'maxima::macsymatoplevel)  as dumpclisp.lisp file and create the memory image, and wrote a short shell script to start maxima with 'maxima' on the CL.  #!/bin/sh exec /local/clisp/bin/clisp norc M /local/maxima/maxima.mem  After that maxima seems to work, i.e. the header and prompt come as expected:  tcsh 1> maxima Maxima 5.4 Mon Jun 26 12:37:38 CDT 2000 (with enhancements by W. Schelter). Licensed under the GNU Public License (see file COPYING) (C1)  and simple things like adding numbers and displaying expressions work. However, most function calls don't work. Our maxima user writes: > It doesn't recognize anything (you can try with a simple "factor(expr);" > or "integrate(f(x),x,x0,x1));". > It doesn't load files for the same reason: "load" appears to it as an > undefined function. > Even the final "quit();" is not understood. I tried to exit with CTRLD and get:  Maxima 5.4 Mon Jun 26 12:37:38 CDT 2000 (with enhancements by W. Schelter). Licensed under the GNU Public License (see file COPYING) (C1) ***  FUNCALL: the function BYE is undefined 1. Break [1]> USER[2]>  Has anybody tried this and can answer some of these questions: 1. Did I do something wrong during the installation (e.g. is there something missing or screwed up in the dumpclisp.lisp file I came up with)? 2. Are the many warnings I get when compiling maxima normal? 3. Sam warned that the patch is work in progress. How much of maxima is known to work (5% or 95%)? 4. Does it look like there's something wrong with the underlying clisp installation? Background: gcl does not install on our platform, which precludes installing the standard maxima dist "out of the box". So, I found out about clisp, which seems to be smaller and faster than gcl (and is known to work on our platform). Any hints and suggestions greatly appreciated. Nikolas 
From: Sam Steingold <sds@gn...>  20000906 19:28:15

Thanks for your patience. I got back to my box and now I can do some more work. > * In message <200008291822.NAA03101@...> > * On the subject of "[clisplist] clisp maxima patch" > * Sent on Tue, 29 Aug 2000 13:22:07 0500 (CDT) > * Honorable Nikolas Kauer <kauer@...> writes: > > this post refers to option 2 of RE: [clisplist] punimax: > > 2. if you would rather use the real thing (maxima), you should > > > > A. visit http://www.ma.utexas.edu/users/wfs/maxima.html > > and get the current CVS version of maxima > > > > B. apply the appended patch (BEWARE: work in progress!) > > > > C. get the latest version of CLISP (http://clisp.cons.org) > > > > D. follow the instructions in maxima/README > > I installed clisp20000306 with the clispalphadecosf4.0d.diff > patch on our Alpha cluster running Tru64 Unix 4.0E, and the test > suite ran fine. I then followed the above instructions and tried > to install maxima for one of our postdocs. The patch worked fine on a > recent CVS version (except for src/homog.lisp, where the indentation > changed slighly). The makefile worked fine, except that the file > src/dumpclisp.lisp was missing. I am sorry  here is the correct file:  ;;; load this to dump maxima under CLISP (load "sysdef.lisp") (inpackage :maxima) (make:make :maxima) (defun bye () (lisp:quit)) (lisp:gc) (lisp:saveinitmem "maxima.mem" :initfunction #'maxima::macsymatoplevel) (lisp:quit)  > and simple things like adding numbers and displaying expressions I get this: $ maxima Maxima 5.4 Mon Jun 26 12:37:38 CDT 2000 (with enhancements by W. Schelter). Licensed under the GNU Public License (see file COPYING) (C1) F(X):=X^2+Y; 2 (D1) F(X) := X + Y (C2) F(2); (D2) Y + 4 (C3) EV(F(2),Y:7); (D3) 11 (C4) F(X):=SIN(X)^2+1; 2 (D4) F(X) := SIN (X) + 1 (C5) F(X+1); 2 (D5) SIN (X + 1) + 1 (C6) G(Y,Z):=F(Z)+3*Y; (D6) G(Y, Z) := F(Z) + 3 Y (C7) EV(G(2*Y+Z,0.5),Y:7); (D7) 3 (Z + 14) + 1.22984884706593 ... > work. However, most function calls don't work. Our maxima user writes: > > It doesn't recognize anything (you can try with a simple "factor(expr);" > > or "integrate(f(x),x,x0,x1));". > > It doesn't load files for the same reason: "load" appears to it as an > > undefined function. > > Even the final "quit();" is not understood. > I tried to exit with CTRLD and get: please use my dumpclisp.lisp. > Has anybody tried this and can answer some of these questions: > 1. Did I do something wrong during the installation (e.g. > is there something missing or screwed up in the dumpclisp.lisp > file I came up with)? yep  use the version in this message. > 2. Are the many warnings I get when compiling maxima normal? yes, for now. I hope to eliminate them eventually. > 3. Sam warned that the patch is work in progress. How much of > maxima is known to work (5% or 95%)? I have no idea yet, but if you send me a test case (your input, what you should get as the output, the output you actually get), we could get farther. > 4. Does it look like there's something wrong with the underlying > clisp installation? no. Thanks for trying my patch. Please do not give up.  Sam Steingold (http://www.podval.org/~sds) Micros**t is not the answer. Micros**t is a question, and the answer is Linux, (http://www.linux.org) the choice of the GNU (http://www.gnu.org) generation. Oh Lord, give me the source code of the Universe and a good debugger! 