From: Javed Alam <jalam@cc...>  20030919 19:46:08

David Thanks for the tips both of the approaches suggested by you worked in getting maxima{compiled with clisp} shell to open up in texmacs {winxp/win200/cygwin/}. However, that is not the end of my problems. I tried to solve this 2nd order differential equation and end up with a wrong answer because if I try doing the same with maxima compiled with GCL I get the right answer. Did some body else had the same problem? $ maxima i i i i i i i ooooo o ooooooo ooooo ooooo I I I I I I I 8 8 8 8 8 o 8 8 I \ `+' / I 8 8 8 8 8 8 \ `+' / 8 8 8 ooooo 8oooo `____' 8 8 8 8 8  8 o 8 8 o 8 8 + ooooo 8oooooo ooo8ooo ooooo 8 Copyright (c) Bruno Haible, Michael Stoll 1992, 1993 Copyright (c) Bruno Haible, Marcus Daniels 19941997 Copyright (c) Bruno Haible, Pierpaolo Bernardi, Sam Steingold 1998 Copyright (c) Bruno Haible, Sam Steingold 19992003 Maxima 5.9.0 http://maxima.sourceforge.net Distributed under the GNU Public License. See the file COPYING. Dedicated to the memory of William Schelter. This is a development version of Maxima. The function bug_report() provides bug reporting information. ;; Loading file /usr/local/share/maxima/5.9.0/share/maximainit.lisp ... ;; Loaded file /usr/local/share/maxima/5.9.0/share/maximainit.lisp (C1) 'diff(x,t,2) + a*x = b*sin(c*t); 2 d x (D1)  + a x = b SIN(c t) 2 dt (C2) ode2(d1,x,t); Is a positive, negative, or zero? positive; / [ (D2) x = b SIN(SQRT(a) t) I (SIN(c t + SQRT(a) t) + SIN(c t  SQRT(a) t)) ] 1. Break [1]> d /(2 COS(SQRT(a) t) ( (SIN(SQRT(a) t))) dt d  2 SIN(SQRT(a) t) ( (COS(SQRT(a) t)))) dt dt / [ + b COS(SQRT(a) t) I (COS(c t + SQRT(a) t)  COS(c t  SQRT(a) t)) ] / d /(2 COS(SQRT(a) t) ( (SIN(SQRT(a) t))) dt d  2 SIN(SQRT(a) t) ( (COS(SQRT(a) t)))) dt + %K1 SIN(SQRT(a) t) dt + %K2 COS(SQRT(a) t) Same solution using maxima compiled under gclisp C:\Program Files\Maxima5.9.0\lib\maxima\5.9.0\binarygcl>maxima eval n) GCL (GNU Common Lisp) Version(2.5.0) Thu Jan 30 02:37:51 CST 2003 Licensed under GNU Library General Public License Contains Enhancements by W. Schelter Use (help) to get some basic information on how to use GCL. Maxima 5.9.0 http://maxima.sourceforge.net Distributed under the GNU Public License. See the file COPYING. Dedicated to the memory of William Schelter. This is a development version of Maxima. The function bug_report() provides bug reporting information. (C3) 'diff(x,t,2) + a*x= b*sin(c*t); 2 d x (D3)  + a x = b SIN(C t) 2 dt (C4) ode2(d3,x,t); Is a positive, negative, or zero? positive; b SIN(C t) (D4) x =   + %K1 SIN(SQRT(a) t) + %K2 COS(SQRT(a) t) 2 C  a (C5) The two solutions are totally different. This has to do with the CLISP and GCLISP compilers. Any help or suggestions in this regard will be highly appreciated. regards javed Billinghurst, David (CALCRTS) wrote: >>From: Billinghurst, David (CALCRTS) >> >> >> >>>From: Javed Alam >>>Sent: Saturday, 13 September 2003 12:02 AM >>> >>>I have been trying to run texmacs1.0.1.21 with maxima5.9.0 under >>>cygwin on a win/xp and win/2000 system. >>> >>>[...] >>> >>>I get the following error. >>> >>>[...] >>> >>>Fatal error: bad url in 'complete' >>>See file : url.cpp >>> >>> >>See if https://savannah.gnu.org/bugs/?func=detailbug&group_id=156&bug_id=4241 >>helps. It suggests hard coding a file location. I will have a look >>myself over weekend. >> >> > >Alternatively: >http://www.mailarchive.com/finkusers@.../msg09820.html >indicates that the problem is related to locate/updatedb. > >When "locate toc_maxima.html" finds the file then texmacs will run maxima. >However, I have some unresolved issues on cygwin: >  the machine that used to work now doesn't, >  the one that didn't now does after doing an updatedb >  updatedb is giving me some grief. > > > >  *Javed Alam, *Ph.D. *Professor* *Civil/Environmental/Chemical Engineering Department* Youngstown State University Youngstown, Ohio 44555 Phone 3309413029 Fax 3309413265 email jalam@... <mailto:jalam@...> web: http://www.eng.ysu.edu/~jalam/ <http://www.eng.ysu.edu/%7Ejalam/>; ====================================================== "Knowledge will forever govern ignorance, and a people who mean to be their own governors, must arm themselves with the power knowledge gives. A popular government without popular information or the means of acquiring it, is but a prologue to a farce or a tragedy or perhaps both."  James Madison /(Fourth President of the United States)/ <http://opengov1.media.mit.edu/>; "Facts are stupid things..."  Ronald Reagan /(40th President of the United States)/ <http://opengov1.media.mit.edu/>; ======================================================== 
From: Richard Fateman <fateman@cs...>  20030919 20:02:04

Usually the way to tell if something is a compiler bug is to find load the corresponding noncompiled file and run the program again. Then try the program again but load only one definition from the file and see if it fixes it. Which one file, which one definition... that's tricky to determine. Another way, if you have 2 different behaviors, is to trace all the programs that you think might matter, and look at the two traces. (output the trace info to text files...) Good luck. RJF Javed Alam wrote: > > David > > Thanks for the tips both of the approaches suggested by you worked in > getting > maxima{compiled with clisp} shell to open up in texmacs > {winxp/win200/cygwin/}. However, that is not the end > of my problems. I tried to solve this 2nd order differential equation > and end up with > a wrong answer because if I try doing the same with maxima compiled > with GCL I get the > right answer. Did some body else had the same problem? > > > > $ maxima > i i i i i i i ooooo o ooooooo ooooo ooooo > I I I I I I I 8 8 8 8 8 o 8 8 > I \ `+' / I 8 8 8 8 8 8 > \ `+' / 8 8 8 ooooo 8oooo > `____' 8 8 8 8 8 >  8 o 8 8 o 8 8 > + ooooo 8oooooo ooo8ooo ooooo 8 > > Copyright (c) Bruno Haible, Michael Stoll 1992, 1993 > Copyright (c) Bruno Haible, Marcus Daniels 19941997 > Copyright (c) Bruno Haible, Pierpaolo Bernardi, Sam Steingold 1998 > Copyright (c) Bruno Haible, Sam Steingold 19992003 > > Maxima 5.9.0 http://maxima.sourceforge.net > Distributed under the GNU Public License. See the file COPYING. > Dedicated to the memory of William Schelter. > This is a development version of Maxima. The function bug_report() > provides bug reporting information. > ;; Loading file /usr/local/share/maxima/5.9.0/share/maximainit.lisp ... > ;; Loaded file /usr/local/share/maxima/5.9.0/share/maximainit.lisp > (C1) 'diff(x,t,2) + a*x = b*sin(c*t); > 2 > d x > (D1)  + a x = b SIN(c t) > 2 > dt > (C2) ode2(d1,x,t); > Is a positive, negative, or zero? > > positive; > / > [ > (D2) x = b SIN(SQRT(a) t) I (SIN(c t + SQRT(a) t) + SIN(c t  SQRT(a) t)) > ] > 1. Break [1]> > > d > /(2 COS(SQRT(a) t) ( (SIN(SQRT(a) t))) > dt > > d >  2 SIN(SQRT(a) t) ( (COS(SQRT(a) t)))) dt > dt > > / > [ > + b COS(SQRT(a) t) I (COS(c t + SQRT(a) t)  COS(c t  SQRT(a) t)) > ] > / > > d > /(2 COS(SQRT(a) t) ( (SIN(SQRT(a) t))) > dt > > d >  2 SIN(SQRT(a) t) ( (COS(SQRT(a) t)))) dt + %K1 SIN(SQRT(a) t) > dt > > + %K2 COS(SQRT(a) t) > > > > > Same solution using maxima compiled under gclisp > > > C:\Program Files\Maxima5.9.0\lib\maxima\5.9.0\binarygcl>maxima eval > n) > GCL (GNU Common Lisp) Version(2.5.0) Thu Jan 30 02:37:51 CST 2003 > Licensed under GNU Library General Public License > Contains Enhancements by W. Schelter > > Use (help) to get some basic information on how to use GCL. > Maxima 5.9.0 http://maxima.sourceforge.net > Distributed under the GNU Public License. See the file COPYING. > Dedicated to the memory of William Schelter. > This is a development version of Maxima. The function bug_report() > provides bug reporting information. > (C3) 'diff(x,t,2) + a*x= b*sin(c*t); > > 2 > d x > (D3)  + a x = b SIN(C t) > 2 > dt > (C4) ode2(d3,x,t); > > Is a positive, negative, or zero? > > positive; > b SIN(C t) > (D4) x =   + %K1 SIN(SQRT(a) t) + %K2 COS(SQRT(a) t) > 2 > C  a > (C5) > > The two solutions are totally different. This has to do with the CLISP > and GCLISP compilers. Any help or suggestions in this regard > will be highly appreciated. > > regards > > javed > > > > > Billinghurst, David (CALCRTS) wrote: > >>> From: Billinghurst, David (CALCRTS) >>> >>> >>>> From: Javed Alam >>>> Sent: Saturday, 13 September 2003 12:02 AM >>>> >>>> I have been trying to run texmacs1.0.1.21 with maxima5.9.0 under >>>> cygwin on a win/xp and win/2000 system. >>>> >>>> [...] >>>> >>>> I get the following error. >>>> >>>> [...] >>>> >>>> Fatal error: bad url in 'complete' >>>> See file : url.cpp >>>> >>> >>> See if >>> https://savannah.gnu.org/bugs/?func=detailbug&group_id=156&bug_id=4241 >>> helps. It suggests hard coding a file location. I will have a look >>> myself over weekend. >>> >> >> >> Alternatively: >> http://www.mailarchive.com/finkusers@.../msg09820.html >> >> indicates that the problem is related to locate/updatedb. >> When "locate toc_maxima.html" finds the file then texmacs will run >> maxima. However, I have some unresolved issues on cygwin: >>  the machine that used to work now doesn't,  the one that didn't >> now does after doing an updatedb >>  updatedb is giving me some grief. >> >> >> >> > > 
From: James Amundson <amundson@us...>  20030919 20:25:06

On Fri, 20030919 at 01:31, Javed Alam wrote: > I tried to solve this 2nd order differential equation > and end up with > a wrong answer because if I try doing the same with maxima compiled with > GCL I get the > right answer. Did some body else had the same problem? <snip> > (C1) 'diff(x,t,2) + a*x = b*sin(c*t); > 2 > d x > (D1)  + a x = b SIN(c t) > 2 > dt > (C2) ode2(d1,x,t); > Is a positive, negative, or zero? > > positive; > / > [ > (D2) x = b SIN(SQRT(a) t) I (SIN(c t + SQRT(a) t) + SIN(c t  SQRT(a) t)) > ] > 1. Break [1]> > > d > /(2 COS(SQRT(a) t) ( (SIN(SQRT(a) t))) > dt > > d >  2 SIN(SQRT(a) t) ( (COS(SQRT(a) t)))) dt > dt > > / > [ > + b COS(SQRT(a) t) I (COS(c t + SQRT(a) t)  COS(c t  SQRT(a) t)) > ] > / > > d > /(2 COS(SQRT(a) t) ( (SIN(SQRT(a) t))) > dt > > d >  2 SIN(SQRT(a) t) ( (COS(SQRT(a) t)))) dt + %K1 SIN(SQRT(a) t) > dt > > + %K2 COS(SQRT(a) t) I was not able to reproduce this result with Maxima 5.9.0+Clisp 2.29 or the current Maxima cvs+Clisp 2.31. Looking at it, I don't see how you could possibly get "1. Break [1]>" In the middle of a displayed equation. Do you know what happened there? It would be helpful for you to send us the output of "build_info();" Jim 
From: Javed Alam <jalam@cc...>  20030919 22:17:49

Hi I ran the maxima versions(5.9.0) compiled with clisp(2.31) and gclisp(2.5.0) for exact same problems. The build_info for both is included. OS ENVIRONMENT: CYGWIN under WINXP CYGWIN { 2003/08/31 12:05:44 Starting cygwin install, version 2.340.2.5 } Here is what I am getting as output: MAXIMA version compiled with the clisp runs most of the problems but fails to produce results for 1. problems involving differentiation of functions 2. 2nd order nonhomogeneous ordinary differential equations. It works ok for 2nd order homogenous ordinary equations. Other than this everything else seems to work including plot2d and plot3d functions. I am including the screen output from both runs : CLISP run jalam@... ~ $ maxima i i i i i i i ooooo o ooooooo ooooo ooooo I I I I I I I 8 8 8 8 8 o 8 8 I \ `+' / I 8 8 8 8 8 8 \ `+' / 8 8 8 ooooo 8oooo `____' 8 8 8 8 8  8 o 8 8 o 8 8 + ooooo 8oooooo ooo8ooo ooooo 8 Copyright (c) Bruno Haible, Michael Stoll 1992, 1993 Copyright (c) Bruno Haible, Marcus Daniels 19941997 Copyright (c) Bruno Haible, Pierpaolo Bernardi, Sam Steingold 1998 Copyright (c) Bruno Haible, Sam Steingold 19992003 Maxima 5.9.0 http://maxima.sourceforge.net Distributed under the GNU Public License. See the file COPYING. Dedicated to the memory of William Schelter. This is a development version of Maxima. The function bug_report() provides bug reporting information. ;; Loading file /usr/local/share/maxima/5.9.0/share/maximainit.lisp ... ;; Loaded file /usr/local/share/maxima/5.9.0/share/maximainit.lisp (C1) build_info; (D1) BUILD_INFO (C2) build_info(); Maxima version: 5.9.0 Maxima build date: 6:33 9/12/2003 host type: i686pccygwin lispimplementationtype: CLISP lispimplementationversion: 2.31 (released 20030901) (built 3272325060) (memo ry 3272351601) (D2) (C3) 2 + 2; (D3) 4 (C4) integrate(x^2,x); 3 x (D4)  3 (C5) integrate(x^3,x,0,1); 1 (D5)  4 (C6) 'diff(x,t,2)=0; 2 d x (D6)  = 0 2 dt (C7) ode2(d6,x,t); (D7) x = %K2 t + %K1 (C8) 'diff(x,t,2) + a*'diff(x,t) + b*x=0; 2 d x dx (D8)  + a  + b x = 0 2 dt dt (C9) ode2(d8,x,t); 2 Is 4 b  a positive, negative, or zero? positive; a t   2 2 2 SQRT(4 b  a ) t SQRT(4 b  a ) t (D9) x = %E (%K1 SIN() + %K2 COS()) 2 2 (C10) solve(x^2 3*x +2=0,x); (D10) [x = 1, x = 2] (C11) diff(x,x); (D11) 1 (C12) diff(x^2,x); d 2 (D12)  (x ) dx (C13) 'diff(x,t,2) +a*'diff(x,t) + b*x =d*sin(e*t); 2 d x dx (D13)  + a  + b x = d SIN(e t) 2 dt dt (C14) ode2(d13,x,t); 2 Is 4 b  a positive, negative, or zero? positive; a t   2 2 SQRT(4 b  a ) t (D14) x = %E (d SIN() 2 / 2 2 [ 2 e t + SQRT(4 b  a ) t SQRT(4 b  a ) t  2 e t I (SIN()  SIN()) ] 2 2 / a t 2   2 SQRT(4 b  a ) t d 2 SQRT(4 b  a ) t /(2 COS() ( (%E SIN())) 2 dt 2 a t 2   2 SQRT(4 b  a ) t d 2 SQRT(4 b  a ) t  2 SIN() ( (%E COS()))) dt 2 dt 2 2 / 2 SQRT(4 b  a ) t [ 2 e t + SQRT(4 b  a ) t + d COS() I (COS() 2 ] 2 / 2 2 SQRT(4 b  a ) t  2 e t SQRT(4 b  a ) t  COS())/(2 COS() 2 2 a t   2 2 d 2 SQRT(4 b  a ) t SQRT(4 b  a ) t ( (%E SIN()))  2 SIN() dt 2 2 a t   2 d 2 SQRT(4 b  a ) t ( (%E COS()))) dt) dt 2 (C15)   2 2 2 SQRT(4 b  a ) t SQRT(4 b  a ) t + %E (%K1 SIN() + %K2 COS()) 2 2 (C15) GCLISP RUN for the sake of comparison jalam@... ~ jalam@... /tmp/maxima/lib/maxima/5.9.0/binarygcl $ ./maxima eval "(user::run)" GCL (GNU Common Lisp) Version(2.5.0) Thu Jan 30 02:37:51 CST 2003 Licensed under GNU Library General Public License Contains Enhancements by W. Schelter Use (help) to get some basic information on how to use GCL. Maxima 5.9.0 http://maxima.sourceforge.net Distributed under the GNU Public License. See the file COPYING. Dedicated to the memory of William Schelter. This is a development version of Maxima. The function bug_report() provides bug reporting information. (C1) build_info; (D1) BUILD_INFO (C2) build_info(); Maxima version: 5.9.0 Maxima build date: 19:10 2/9/2003 host type: i686pcmingw32 lispimplementationtype: Kyoto Common Lisp lispimplementationversion: GCL25.0 (D2) (C3) 2 + 2; (D3) 4 (C4) integrate (x^2,x); 3 x (D4)  3 (C5) integrate(x^3,x,0,1); 1 (D5)  4 (C6) 'diff(x,t,2)=0; 2 d x (D6)  = 0 2 dt (C7) ode2(d6,x,t); (D7) x = %K2 t + %K1 (C8) 'diff(x,t,2) + a*'diff(x,t) + b*x=0; 2 d x dx (D8)  + a  + b x = 0 2 dt dt (C9) ode2(d8,x,t); 2 Is 4 b  a positive, negative, or zero? positive; a t   2 2 2 SQRT(4 b  a ) t SQRT(4 b  a ) t (D9) x = %E (%K1 SIN() + %K2 COS()) 2 2 (C10) solve(x^2  3*x +2=0,x); (D10) [x = 1, x = 2] (C11) diff(x,x); (D11) 1 (C12) diff(x^2,x); (D12) 2 x (C13) 'diff(x,t,2) + a*'diff(x,t) + b*x = d*sin(e*t); 2 d x dx (D13)  + a  + b x = d SIN(e t) 2 dt dt (C14) ode2(d13,x,t); 2 Is 4 b  a positive, negative, or zero? positive; a t   2 2 2 SQRT(4 b  a ) t SQRT(4 b  a ) t (D14) x = %E (%K1 SIN() + %K2 COS()) 2 2 2 (d e  b d) SIN(e t) + a d e COS(e t)   4 2 2 2 e + (a  2 b) e + b (C15) This shows that it could be a very small problem in clisp compilation of maxima on winxp under cygwin. Any suggestion will be very helpful. Thanks regards javed James Amundson wrote: >On Fri, 20030919 at 01:31, Javed Alam wrote: > > >>I tried to solve this 2nd order differential equation >>and end up with >>a wrong answer because if I try doing the same with maxima compiled with >>GCL I get the >>right answer. Did some body else had the same problem? >> >> > ><snip> > > > >>(C1) 'diff(x,t,2) + a*x = b*sin(c*t); >> 2 >> d x >>(D1)  + a x = b SIN(c t) >> 2 >> dt >>(C2) ode2(d1,x,t); >>Is a positive, negative, or zero? >> >>positive; >> / >> [ >>(D2) x = b SIN(SQRT(a) t) I (SIN(c t + SQRT(a) t) + SIN(c t  SQRT(a) t)) >> ] >>1. Break [1]> >> >> d >>/(2 COS(SQRT(a) t) ( (SIN(SQRT(a) t))) >> dt >> >> d >>  2 SIN(SQRT(a) t) ( (COS(SQRT(a) t)))) dt >> dt >> >> / >> [ >> + b COS(SQRT(a) t) I (COS(c t + SQRT(a) t)  COS(c t  SQRT(a) t)) >> ] >> / >> >> d >>/(2 COS(SQRT(a) t) ( (SIN(SQRT(a) t))) >> dt >> >> d >>  2 SIN(SQRT(a) t) ( (COS(SQRT(a) t)))) dt + %K1 SIN(SQRT(a) t) >> dt >> >> + %K2 COS(SQRT(a) t) >> >> > >I was not able to reproduce this result with Maxima 5.9.0+Clisp 2.29 or >the current Maxima cvs+Clisp 2.31. Looking at it, I don't see how you >could possibly get "1. Break [1]>" In the middle of a displayed >equation. Do you know what happened there? It would be helpful for you >to send us the output of "build_info();" > >Jim > > > > >  *Javed Alam, *Ph.D. *Professor* *Civil/Environmental/Chemical Engineering Department* Youngstown State University Youngstown, Ohio 44555 Phone 3309413029 Fax 3309413265 email jalam@... <mailto:jalam@...> web: http://www.eng.ysu.edu/~jalam/ <http://www.eng.ysu.edu/%7Ejalam/>; ====================================================== "Knowledge will forever govern ignorance, and a people who mean to be their own governors, must arm themselves with the power knowledge gives. A popular government without popular information or the means of acquiring it, is but a prologue to a farce or a tragedy or perhaps both."  James Madison /(Fourth President of the United States)/ <http://opengov1.media.mit.edu/>; "Facts are stupid things..."  Ronald Reagan /(40th President of the United States)/ <http://opengov1.media.mit.edu/>; ======================================================== 
From: James Amundson <amundson@us...>  20030920 00:49:23

On Fri, 20030919 at 17:09, Javed Alam wrote: > Hi > > I ran the maxima versions(5.9.0) compiled with clisp(2.31) and > gclisp(2.5.0) for exact same problems. The build_info for both is included. > OS ENVIRONMENT: CYGWIN under WINXP > CYGWIN { 2003/08/31 12:05:44 Starting cygwin install, version 2.340.2.5 } I am pretty sure there is a small, but important conflict between clisp 2.31 and Maxima 5.9.0. The good news is that the problem is fixed in CVS. The bad news is that I don't remember exactly which change it was. (Anyone else? It may have been a change for CMUCL. It definitely predates Clisp 2.31.) We need to think about making a 5.9.1 release. In the meantime, getting the CVS version of Maxima would be the easiest workaround. Jim 
From: Fred Cohen <fc@al...>  20030920 01:24:20

So we are all clear, which versions have this problem? And on another related issue  how can it really be that we get a working functional system with that wrong an answer? It dramtically reduces my confidence in the tests and related material associated with maxima / clisp if such problems can happen so easily without notice. FC Per the message sent by James Amundson: > On Fri, 20030919 at 17:09, Javed Alam wrote: > > Hi > > > > I ran the maxima versions(5.9.0) compiled with clisp(2.31) and > > gclisp(2.5.0) for exact same problems. The build_info for both is included. > > OS ENVIRONMENT: CYGWIN under WINXP > > CYGWIN { 2003/08/31 12:05:44 Starting cygwin install, version 2.340.2.5 } > I am pretty sure there is a small, but important conflict between clisp > 2.31 and Maxima 5.9.0. The good news is that the problem is fixed in > CVS. The bad news is that I don't remember exactly which change it was. > (Anyone else? It may have been a change for CMUCL. It definitely > predates Clisp 2.31.) > We need to think about making a 5.9.1 release. In the meantime, getting > the CVS version of Maxima would be the easiest workaround. > Jim >  > This sf.net email is sponsored by:ThinkGeek > Welcome to geek heaven. > http://thinkgeek.com/sf > _______________________________________________ > clisplist mailing list > clisplist@... > https://lists.sourceforge.net/lists/listinfo/clisplist  This communication is confidential to the parties it is intended to serve  Fred Cohen  http://all.net/  fc@...  fc@...  tel/fax: 9254540171 Fred Cohen & Associates  University of New Haven  Security Posture 
From: James Amundson <amundson@us...>  20030921 01:20:39

On Fri, 19 Sep 2003 18:23:47 0700, Fred Cohen wrote: > So we are all clear, which versions have this problem? Clisp 2.29 works well with Maxima 5.9.0. Clisp 2.30 has some problem with numerical type conversions that causes huge problems with Maxima, as is documented in the file README.lisps, which is included in the Maxima distribution.. Clisp 2.31 has "this problem" with Maxima 5.9.0, but the problem is fixed in the current Maxima cvs. > And on another related issue  how can it really be that we get a > working functional system with that wrong an answer? Maxima comes with a test suite, which can be invoked with "make check" in the source directory. When compiled with Clisp 2.31, Maxima 5.9.0 crashes roughly 1/3 of the way into the test suite. It isn't a working, functional system. > It dramtically > reduces my confidence in the tests and related material associated with > maxima / clisp if such problems can happen so easily without notice. The current Maxima test suite easily identifies the problem. I would be worried if it didn't. Jim Amundson 
From: Vadim V. Zhytnikov <vvzhy@ma...>  20030920 04:11:57

James Amundson =D0=C9=DB=C5=D4: > On Fri, 20030919 at 17:09, Javed Alam wrote: >=20 >>Hi >> >>I ran the maxima versions(5.9.0) compiled with clisp(2.31) and=20 >>gclisp(2.5.0) for exact same problems. The build_info for both is inclu= ded. >>OS ENVIRONMENT: CYGWIN under WINXP >>CYGWIN { 2003/08/31 12:05:44 Starting cygwin install, version 2.340.2.5= } >=20 >=20 > I am pretty sure there is a small, but important conflict between clisp > 2.31 and Maxima 5.9.0. The good news is that the problem is fixed in > CVS. The bad news is that I don't remember exactly which change it was. > (Anyone else? It may have been a change for CMUCL. It definitely > predates Clisp 2.31.)=20 >=20 > We need to think about making a 5.9.1 release. In the meantime, getting > the CVS version of Maxima would be the easiest workaround. >=20 Right, Maxima 5.9.0 will not work with clisp 2.31 and 2.30 (other=20 reason). As for clisp 2.31 incompatibolity  the fix is the latest patch to commac.lisp (maclisptypep). =20 Vadim V. Zhytnikov <vvzhy@...> <vvzhy@...> 
From: James Amundson <amundson@us...>  20030922 15:22:46

On Mon, 20030922 at 09:24, Sam Steingold wrote: > The following message is a courtesy copy of an article > that has been posted to gmane.lisp.clisp.general as well. > > > * James Amundson <nzhaqfba@...> [20030920 19:49:52 0500]: > > When compiled with Clisp 2.31, Maxima 5.9.0 crashes roughly 1/3 of the > > way into the test suite. It isn't a working, functional system. > > could you please send us with a test case? If you have a maxima source directory, just type "make check". You should get "Unkown error in log file." The details can be seen in tests/testsclisp.log. The problem is completely maxima's fault. Raymond Toy fixed the problem in CVS a few months ago. I've attached the relevant section of code below. (The buggy function is the one that now has #+nil.) We will try to make a new release soon. Jim Relevant section of commac.lisp: ;; This assumes way too much about how typep works, so don't use it. ;; We leave it here for reference to make sure the replacement below ;; does the same thing. #+nil (defun maclisptypep (x &optional type) (cond (type (lisp:let (( pred (get type 'mltypep))) (cond (pred (funcall pred x)) (t (typep x type))))) (t (lisp:let ((.type. (#. (if (boundp '*primitivedatatypefunction*) *primitivedatatypefunction* 'typeof) x))) (cond ((oneoftypes .type. 'hi nil) 'symbol) ((oneoftypes .type. '(a)) 'list) ((oneoftypes .type. 3) 'fixnum) ((oneoftypes .type. (makearray 3) "abc") (cond ((stringp x) 'string) ;;should really be symbol 'ugggh #+ti ((hashtablep x) 'hashtable) (t 'array))) ((oneoftypes .type. (expt 2 50) 1.234 mostpositivesinglefloat mostpositivedoublefloat mostpositivelongfloat ) (cond ((integerp x) 'bignum) ((floatp x) 'flonum ) (t 'number))) ;;note the following is 'random in maclisp ((oneoftypes .type. #'cons) 'compiledfunction) #ti ((oneoftypes .type. (makehashtable)) (cond ((hashtablep x) 'hashtable) (t (typeof x)))) ((arrayp x) 'array) ;((oneoftypes .type. (makearray '(2 3))) 'array) (t (typeof x))))))) ;; A more portable implementation of maclisptypep. I (rtoy) think it ;; would probably be better to replace uses of maclisptypep and/or ;; mltypep with the corresponding Common Lisp typep or typeof or ;; subtypep, as appropriate. (defun maclisptypep (x &optional type) (cond (type (lisp:let ((pred (get type 'mltypep))) (cond (pred (funcall pred x)) (t (typep x type))))) (t (typecase x (cl:cons 'list) (cl:fixnum 'fixnum) (cl:integer 'bignum) (cl:float 'flonum) (cl:number 'number) (cl:array 'array) (cl:hashtable 'hashtable) (t (typeof x)))))) 
From: James Amundson <amundson@us...>  20030922 17:50:57

On Mon, 20030922 at 10:45, Sam Steingold wrote: > thanks  I thought that "crash" meant a CLISP sergfault (which is always > a CLISP fault, not an application problem). That's a reasonable assumption. What I really meant is that Maxima falls into the Lisp debugger, which is Maxima's equivalent of crashing. Sorry for causing unnecessary alarm. Jim 