[Maxima-bugs] [ maxima-Bugs-3111568 ] subsequent calls to gradef hide variable dependencies

[SourceForge.net <no...@so...>]

Bugs item #3111568, was opened at 2010-11-18 15:09
Message generated for change (Comment added) made by tnfnk
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=104933&aid=3111568&group_id=4933

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Lisp Core
Group: None
Status: Closed
Resolution: None
Priority: 5
Private: No
Submitted By: tnfnk (tnfnk)
Assigned to: Nobody/Anonymous (nobody)
Summary: subsequent calls to gradef hide variable dependencies

Initial Comment:
/*
    2010-11-18, T. Finke

    Bug Report: subsequent calls to gradef hide variable dependencies
 
    Maxima version: 5.22.1
    Maxima build date: 14:32 9/19/2010
    Host type: x86_64-unknown-linux-gnu
    Lisp implementation type: CLISP
    Lisp implementation version: 2.44.1 (2008-02-23) (built on build35 [127.0.0.1])

    The Maxima manual states for the System variable "dependencies": 

    "... The dependencies list is cumulative: 
         each call to depends or gradef appends additional items. "

    The following sequence shows, that subsequent calls to "gradef()"
    concerning variables will hide previously declared
    dependencies. Only the most recently declared dependency is listed
    by "dependencies". But the "diff()" operator regards former
    declarations if called exactly for the dpendent variable. If
    called for expressions, former dependencies are ignored, yielding
    zero results. 

    If the dependency later is declared by "depends" for all
    independents, further calculations will be correct. 

    It seems that "gradef" - if called for variables rather than for
    functions - fails to add subsequent dependencies. 

*/

gradef(f, x, g);
gradef(f, y, h);
diff(f, x);         /* expect "g", correct */
diff(f, y);         /* expect "h", correct */
diff(c * f, y);     /* expect "c h", correct */
diff(c * f, x);     /* expect "c g", but get "0" -- ERROR */
dependencies;       /* expect "[f(x, y)]", but get "[f(y)]" -- WRONG */
depends(f, [x, y]);
diff(c * f, y);     /* expect "c h", again correct, as above */
diff(c * f, x);     /* expect "c g", NOW correct */
dependencies;       /* expect "[f(x, y)]", NOW correct */



----------------------------------------------------------------------

Comment By: tnfnk (tnfnk)
Date: 2010-11-21 03:09

Message:
Thank you very much for the information about the fix.

May I suggest to modify the documentation about "depends", "dependencies"
and also "gradef" concerning subsequent calls. Also it may be useful to
mention that a call to "depends" should follow the call of "gradef"
(hopefully I am correct with this). 

Thanks again - especially for your great work on Maxima

Torsten Finke


----------------------------------------------------------------------

Comment By: Dieter Kaiser (crategus)
Date: 2010-11-18 22:07

Message:
Thank you very much for this bug report:
Fixed in comm.lisp revision 1.56.
The functions depends has changed and does no longer append new
dependencies. This behavior is restored for the function gradef.
Closing this bug report as fixed.
Dieter Kaiser

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=104933&aid=3111568&group_id=4933