maxima-bugs — Reports from the SF bug-tracking system

[Maxima-bugs] [maxima:bugs] #4726 solve doesn't really solve

[Stavros M. <mac...@us...>]

> Anyway i found something where i had an error in assume(n>2,m,constant) should have put assume(n>2,M,constant) and when i did that it did solve directly for y. 

No, if you look at the output of 

assume(n>2,M,constant)

you will see that the "M,constant" part had no effect:

  =>   [n > 2, meaningless, meaningless]

You presumably want declare(M,constant).



---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Sun May 03, 2026 10:44 PM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] Re: #4726 solve doesn't really solve

[Stavros M. <mac...@us...>]

I meant that you wrote

(assume_pos:true, file_output_append:true, ratprint:false, showtime:true,
load(simplify_sum), intanalysis: false, simpsum:true, load("lrats")
,letrat:true, ratfac:true,algebraic:true,
rootsconmode=super,algexact:true,fpprintprec:4);


which means that the individual results from the statements weren't visible
to you when you ran it.

If they had been visible, you would presumably have noticed that*
rootsconmode=super * returns *rootsconmode=super*, and not *super*, which
is what it would have returned if you had successfully assigned to the
variable *rootsconmode*. Just a typo (after all, you used the correct
syntax for all the other globals you set).

Then you wrote

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));


Again, the individual statements' output wasn't visible to you. If it had
been, you would presumably have noticed that

(assume(n>2,m,constant)


returns

[(n+(-2))+2 > 2,meaningless,meaningless]


which is a signal that *assume(m,constant)* isn't the correct syntax.

               -s

On Sun, May 3, 2026 at 6:44 PM dan hayes <zm...@us...> wrote:

> Re above remarks by Stavros Macrakis , Barton Willis. explain more what
> you mean by
>
> 'One of the disadvantages of the (...) approach to setting multiple
> options is that you don't see any errors.'
>
> 'For clarity, unless it is necessary to use a non-default value for an
> option variable, I suggest using the default values--it makes an example
> easier for me to understand. ' I don't understand exactly what you mean eg
> where did i use non-default value for option and did u try it with the
> default value and did it change things.
> If you are referring to anything in
>
> (assume_pos:true, file_output_append:true, ratprint:false,
> showtime:true,load(simplify_sum), intanalysis: false, simpsum:true,
> load("lrats"),letrat:true, ratfac:true, algebraic:true,
> rootsconmode=super,algexact:true,fpprintprec:4);
>
> That is just what i put in for the configuration to run everytime in top
> menu, Edit,configure "Maxima commands to be executed every time wxMaxima
> starts Maxima" I did not put that in pertaining to this particular problem
> but those are what i generally use for most all my programs and i did not
> think it would have any negative effect here.
>
> Anyway i found something where i had an error in assume(n>2,m,constant)
> should have put assume(n>2,M,constant) and when i did that it did solve
> directly for y. Still i can't see why even if i had not done that it should
> have been equivalent to the longer drawn out way i did it except for the
> multivalued considerations. I think if had put in a constant real positive
> integer value for n then it should have given all the n+1 multivalued
> answers as a trial with n=3 did give 4 answers for y in terms of x.
>
> Also above delete where i wrote ' ungodly difficult problem in finding the
> correct constant of integration to get the answer in this form. ' Because
> it is not so difficult but rather straightforward.
> ------------------------------
>
> *[bugs:#4726] <https://sourceforge.net/p/maxima/bugs/4726/> solve doesn't
> really solve*
>
> *Status:* open
> *Group:* None
> *Created:* Wed Apr 29, 2026 09:28 AM UTC by dan hayes
> *Last Updated:* Thu Apr 30, 2026 11:11 PM UTC
> *Owner:* nobody
>
> WxMaxima version: 26.01.0_MSW
> Using wxWidgets version: wxWidgets 3.2.9
> Maxima version: 5.49.0
> Maxima build date: 2026-01-02 21:27:51
> Host type: x86_64-w64-mingw32
> System type: Win32 10.0.19041 X86-64
> Lisp implementation type: SBCL
> Lisp implementation version: 2.6.0
>
> (assume_pos:true, file_output_append:true, ratprint:false, showtime:true,
> load(simplify_sum), intanalysis: false, simpsum:true, load("lrats")
> ,letrat:true, ratfac:true,algebraic:true,
> rootsconmode=super,algexact:true,fpprintprec:4);
>
>
> (assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));
>
> Now from the output rs should be in the form y=some function of x but it
> isn't so i had to do
>
> [p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];
>
> to get the correct answer. I tried even using solveradcan:true , which was
> default false and even that
> did not even help to get maxima to give the correct answer. I shouldn't
> actually say "the correct"
> answer because there is an arbitrary constant of integration that could be
> used for expression rx in
> rx:M*integrate(1/(1-x/a),x). Since actually i desire the solution such
> that y=0 when x=0 and a text
> then by some mysterious way to get the answer in the form
>
> ["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];
>
> Now i would not necessarily expect maxima to be able to get this final
> form which is an ungodly
> difficult problem in finding the correct constant of integration to get
> the answer in this form but
> it would be nice. Though if anyone knows how to find that constant of
> integration would be nice. I
> also add that many times prior i have had to do it myself rationalizing
> etc. etc. to get the answer from
> maximas unwillingness to fully solve a straightforward problem. This is
> also a feature request to get
> maxima to fully solve the problem rather than just giving an incomplete
> answer.
> ------------------------------
>
> Sent from sourceforge.net because you indicated interest in
> https://sourceforge.net/p/maxima/bugs/4726/
>
> To unsubscribe from further messages, please visit
> https://sourceforge.net/auth/subscriptions/
>



---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Sun May 03, 2026 10:44 PM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4726 solve doesn't really solve

[dan h. <zm...@us...>]

 Re above remarks by Stavros Macrakis , Barton Willis. explain more what you mean by

'One of the disadvantages of the (...) approach to setting multiple options is that you don't see any errors.'

'For clarity, unless it is necessary to use a non-default value for an option variable, I suggest using the default values--it makes an example easier for me to understand. '  I don't understand exactly what you mean eg where did i use non-default value for option and did u try it with the default value and did it change things.
If you are referring to anything in 

(assume_pos:true, file_output_append:true, ratprint:false, showtime:true,load(simplify_sum), intanalysis: false, simpsum:true, load("lrats"),letrat:true,  ratfac:true, algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

That is just what i put in for the configuration to run everytime in top menu, Edit,configure "Maxima commands to be executed every time wxMaxima starts Maxima"  I did not put that in pertaining to this particular problem but those are what i generally use for most all my programs and i did not think it would have any negative effect here.

Anyway i found something where i had an error in  assume(n>2,m,constant) should have put assume(n>2,M,constant) and when i did that it did solve directly for y. Still i can't see why even if i had not done that it should have been equivalent to the longer drawn out way i did it except for the multivalued considerations. I think if had put in a constant real positive integer value for n then it should have given all the n+1 multivalued answers as a trial with n=3 did give 4 answers for y in terms of x.

Also above delete where i wrote ' ungodly difficult problem in finding the correct constant of integration to get the answer in this form. '   Because it is not so difficult but rather straightforward.


---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Thu Apr 30, 2026 11:11 PM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4724 xreduce with init arg and declared nary function

[Kris K. <kj...@us...>]

- **status**: open --> closed
- **Comment**:

Fixed by commit [e4af41].  Closing this ticket.



---

**[bugs:#4724] xreduce with init arg and declared nary function**

**Status:** closed
**Group:** None
**Created:** Sat Apr 25, 2026 05:46 PM UTC by Kris Katterjohn
**Last Updated:** Sat May 02, 2026 07:43 PM UTC
**Owner:** Kris Katterjohn


Users can `declare` functions to be nary; below I'll call these "just-`declare`d-nary".  `xreduce` also handles certain nary cases specially (like `and`, `or`, `append`, ...); below I'll call these "special-cased-nary".  (I'm too lazy to come up with less silly names.)

`xreduce` does not pass a given init arg to just-`declare`d-nary functions, but it does for special-cased-nary functions.

In the examples below `foo` is not nary, `fnary` is just-`declare`d-nary, and (the arbitrarily chosen) `and` is special-cased-nary.

~~~
(%i1) declare (fnary, nary)$

(%i2) xreduce (foo, [], 0); /* correct: the 0 is just returned */
(%o2) 0

(%i3) xreduce (fnary, [], 0); /* wrong: should be fnary(0) */
(%o3) fnary()

(%i4) xreduce ("and", [], 0); /* correct: this is from "and"(0) */
(%o4) 0

(%i5) xreduce (foo, [1, 2, 3], 0); /* correct */
(%o5) foo(foo(foo(0,1),2),3)

(%i6) xreduce (fnary, [1, 2, 3], 0); /* wrong: should be fnary(0,1,2,3) */
(%o6) fnary(1,2,3)

(%i7) xreduce ("and", [1, 2, 3], 0); /* correct */
(%o7) 0 and 1 and 2 and 3
~~~

Regarding the unary cases `%o3` and `%o4`, compare to:

~~~
(%i8) xreduce (fnary, [0]); /* this one is correct */
(%o8) fnary(0)

(%i9) xreduce ("and", [0]); /* again correct: this is also from "and"(0) */
(%o9) 0
~~~

While running the test suite (with share tests), it doesn't seem that `xreduce` is ever called with both an init arg and a just-`declare`d-nary function.

A quick look suggests that this bug was introduced in commit [5557be].

Patch below.  I tried to keep with the style of the surrounding code.

~~~
diff --git a/src/nset.lisp b/src/nset.lisp
index d9970b8ed..05ed8f2d9 100644
--- a/src/nset.lisp
+++ b/src/nset.lisp
@@ -1171,8 +1171,10 @@
 	     (funcall opfn s)))
 
 	  (op-props
-	   ($apply f ($listify s)))
-	  
+	   (setq s (require-list-or-set s '$xreduce))
+	   (unless (eq init 'no-init)
+	     (setq s (cons init s)))
+	   ($apply f (cons '(mlist) s)))
 	  (t
 	   (rl-reduce f ($listify s) nil init '$xreduce)))))
~~~

I'll commit this patch along with some tests if there are no objections.


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] Re: #4724 xreduce with init arg and declared nary function

[Kris K. <kj...@us...>]

Originally (see commit [661c3a]) nset started giving the `nary` property to `+`, `*`, `and`, `or`, `max`, `min`, and `append`.  (Later nset also started giving the `nary` property to `union`. And to_poly_solve also started giving the `nary` property to `%and` and `%or`.)

My guess is that it was not intentional to leave out `sconcat` and specific others.

I'll be closing this ticket a little later today.  Can you open a separate ticket for giving the `nary` property to other stuff?


---

**[bugs:#4724] xreduce with init arg and declared nary function**

**Status:** open
**Group:** None
**Created:** Sat Apr 25, 2026 05:46 PM UTC by Kris Katterjohn
**Last Updated:** Mon Apr 27, 2026 03:15 PM UTC
**Owner:** Kris Katterjohn


Users can `declare` functions to be nary; below I'll call these "just-`declare`d-nary".  `xreduce` also handles certain nary cases specially (like `and`, `or`, `append`, ...); below I'll call these "special-cased-nary".  (I'm too lazy to come up with less silly names.)

`xreduce` does not pass a given init arg to just-`declare`d-nary functions, but it does for special-cased-nary functions.

In the examples below `foo` is not nary, `fnary` is just-`declare`d-nary, and (the arbitrarily chosen) `and` is special-cased-nary.

~~~
(%i1) declare (fnary, nary)$

(%i2) xreduce (foo, [], 0); /* correct: the 0 is just returned */
(%o2) 0

(%i3) xreduce (fnary, [], 0); /* wrong: should be fnary(0) */
(%o3) fnary()

(%i4) xreduce ("and", [], 0); /* correct: this is from "and"(0) */
(%o4) 0

(%i5) xreduce (foo, [1, 2, 3], 0); /* correct */
(%o5) foo(foo(foo(0,1),2),3)

(%i6) xreduce (fnary, [1, 2, 3], 0); /* wrong: should be fnary(0,1,2,3) */
(%o6) fnary(1,2,3)

(%i7) xreduce ("and", [1, 2, 3], 0); /* correct */
(%o7) 0 and 1 and 2 and 3
~~~

Regarding the unary cases `%o3` and `%o4`, compare to:

~~~
(%i8) xreduce (fnary, [0]); /* this one is correct */
(%o8) fnary(0)

(%i9) xreduce ("and", [0]); /* again correct: this is also from "and"(0) */
(%o9) 0
~~~

While running the test suite (with share tests), it doesn't seem that `xreduce` is ever called with both an init arg and a just-`declare`d-nary function.

A quick look suggests that this bug was introduced in commit [5557be].

Patch below.  I tried to keep with the style of the surrounding code.

~~~
diff --git a/src/nset.lisp b/src/nset.lisp
index d9970b8ed..05ed8f2d9 100644
--- a/src/nset.lisp
+++ b/src/nset.lisp
@@ -1171,8 +1171,10 @@
 	     (funcall opfn s)))
 
 	  (op-props
-	   ($apply f ($listify s)))
-	  
+	   (setq s (require-list-or-set s '$xreduce))
+	   (unless (eq init 'no-init)
+	     (setq s (cons init s)))
+	   ($apply f (cons '(mlist) s)))
 	  (t
 	   (rl-reduce f ($listify s) nil init '$xreduce)))))
~~~

I'll commit this patch along with some tests if there are no objections.


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4729 rtest_limit.mac: Problem 230 (line 866)

[Barton W. <wil...@us...>]

- **status**: open --> closed
- **Comment**:

Suggested fix applied in  Commit [74ac5c] master. Closing ticket as closed.



---

**[bugs:#4729] rtest\_limit.mac: Problem 230 \(line 866\) **

**Status:** closed
**Group:** None
**Created:** Sat May 02, 2026 06:32 AM UTC by Barton Willis
**Last Updated:** Sat May 02, 2026 06:32 AM UTC
**Owner:** nobody


The expected result from  rtest_limit.mac: Problem 230 (line 866) is not as accurate as it should be. Consider
~~~
(%i1)	xxx : block([numer : true,ratprint : false], integrate(x^3/(exp(x)-1),x,0,inf));
(xxx)	6.49393940226683

(%i2)	yyy : integrate(x^3/(exp(x)-1),x,0,inf);
(yyy)	%pi^4/15

(%i3)	float((xxx-yyy)/yyy);
(%o3)	2.7354071686905186*10^-16

(%i4)	float(yyy);
(%o4)	6.493939402266828
~~~

A putative fix is to locally set `$float` to `nil` in `intsubs`:

~~~
(defun intsubs (e a b ivar)
  (let (($float nil) ;; NEW!
        (edges (cond ((not $intanalysis)
		      '$no)		;don't do any checking.
		    (t (discontinuities-in-interval 
			(let (($algebraic t)) 
			  (sratsimp e))
			ivar a b)))))
~~~


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4729 rtest_limit.mac: Problem 230 (line 866)

[Barton W. <wil...@us...>]

---

**[bugs:#4729] rtest\_limit.mac: Problem 230 \(line 866\) **

**Status:** open
**Group:** None
**Created:** Sat May 02, 2026 06:32 AM UTC by Barton Willis
**Last Updated:** Sat May 02, 2026 06:32 AM UTC
**Owner:** nobody


The expected result from  rtest_limit.mac: Problem 230 (line 866) is not as accurate as it should be. Consider
~~~
(%i1)	xxx : block([numer : true,ratprint : false], integrate(x^3/(exp(x)-1),x,0,inf));
(xxx)	6.49393940226683

(%i2)	yyy : integrate(x^3/(exp(x)-1),x,0,inf);
(yyy)	%pi^4/15

(%i3)	float((xxx-yyy)/yyy);
(%o3)	2.7354071686905186*10^-16

(%i4)	float(yyy);
(%o4)	6.493939402266828
~~~

A putative fix is to locally set `$float` to `nil` in `intsubs`:

~~~
(defun intsubs (e a b ivar)
  (let (($float nil) ;; NEW!
        (edges (cond ((not $intanalysis)
		      '$no)		;don't do any checking.
		    (t (discontinuities-in-interval 
			(let (($algebraic t)) 
			  (sratsimp e))
			ivar a b)))))
~~~


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4727 - is converted to +- (in a special case)

[Stavros M. <mac...@us...>]

- **status**: open --> not-a-bug



---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** not-a-bug
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Fri May 01, 2026 10:07 PM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] Re: #4727 - is converted to +- (in a special case)

[Stavros M. <mac...@us...>]

Re:
> The question is why
> L(h):=1+1-1h;
> is not simplified to
> L(h) := 2-h

As I said before, the body of a function definition is *never* simplified until the function is called. If you want an expression in the body of a function to be simplified, you can always do something like this:
~~~
L(h):= ''(1+1-1*h);
gives L(h):=2-h
~~~
That evaluates and simplifies the ``''(...)`` expression in the *global* context, which may or may not be what you want. If you want to only simplify and not evaluate, you can write ``''('(...))``.

But why do you care?

I am closing this as not a bug.


---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Thu Apr 30, 2026 07:30 PM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4719 Undefined limit product $INF * $ZEROA in lim-times

[Barton W. <wil...@us...>]

A recent build gives
~~~
(%i2) display2d : false$

(%i3) qqq : %e^(%e^(1/x)*gamma(x-%e^(-x)))-%e^gamma(x);

(%o3) %e^(%e^(1/x)*gamma(x-%e^-x))-%e^gamma(x)
(%i4) limit(qqq,x,inf);

(%o4) -(%e^-1*(%e^2-sqrt(%e)))
~~~
It's  nice that we don't get thrown into the debugger, but the answer is wrong--I'm pretty sure that the correct value is `infinity`. 


---

**[bugs:#4719] Undefined limit product $INF \* $ZEROA in lim-times**

**Status:** open
**Group:** None
**Labels:** limit taylor 
**Created:** Sun Apr 19, 2026 05:29 PM UTC by Barton Willis
**Last Updated:** Sun Apr 19, 2026 05:29 PM UTC
**Owner:** nobody


I have seen this Taylor bug before:

~~~
(%i2)	qqq : %e^(%e^(1/x)*gamma(x-%e^(-x)))-%e^gamma(x);
(%o2) %e^(%e^(1/x)*gamma(x-%e^-x))-%e^gamma(x)
(%i3)	limit(qqq,x,inf);
debugger invoked on a SIMPLE-CONDITION in thread

#<THREAD"main thread" RUNNING {1007DC0183}>:

  Undefined limit product $INF * $ZEROA in lim-times

0] 
~~~


---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4728 Building German documentation fails with texinfo-7.3

[Ulrich M. <uda...@us...>]

---

**[bugs:#4728] Building German documentation fails with texinfo-7.3**

**Status:** open
**Group:** None
**Created:** Fri May 01, 2026 08:42 AM UTC by Ulrich Müller
**Last Updated:** Fri May 01, 2026 08:42 AM UTC
**Owner:** nobody
**Attachments:**

- [maxima-5.49.0-info-de-Makefile.patch](https://sourceforge.net/p/maxima/bugs/4728/attachment/maxima-5.49.0-info-de-Makefile.patch) (282 Bytes; text/x-patch)


Forwarding Gentoo Linux bug https://bugs.gentoo.org/973233:

```
Making all in doc
make[1]: Entering directory '/tmp/portage/sci-mathematics/maxima-5.49.0/work/maxima-5.49.0/doc'
Making all in info
make[2]: Entering directory '/tmp/portage/sci-mathematics/maxima-5.49.0/work/maxima-5.49.0/doc/info'
Making all in de
make[3]: Entering directory '/tmp/portage/sci-mathematics/maxima-5.49.0/work/maxima-5.49.0/doc/info/de'
restore=: && backupdir=".am$$" && \
am__cwd=`pwd` && CDPATH="${ZSH_VERSION+.}:" && cd . && \
rm -rf $backupdir && mkdir $backupdir && \
if (/usr/bin/makeinfo --version) >/dev/null 2>&1; then \
  for f in maxima.info maxima.info-[0-9] maxima.info-[0-9][0-9] maxima.i[0-9] maxima.i[0-9][0-9]; do \
    if test -f $f; then mv $f $backupdir; restore=mv; else :; fi; \
  done; \
else :; fi && \
cd "$am__cwd"; \
if /usr/bin/makeinfo  --enable-encoding -I . -I .. -I ../.. ./.. -I . \
 -o maxima.info maxima.texi; \
then \
  rc=0; \
  CDPATH="${ZSH_VERSION+.}:" && cd .; \
else \
  rc=$?; \
  CDPATH="${ZSH_VERSION+.}:" && cd . && \
  $restore $backupdir/* `echo "./maxima.info" | sed 's|[^/]*$||'`; \
fi; \
rm -rf $backupdir; exit $rc
Encoding name should not be undef at /usr/share/texi2any/Texinfo/ParserNonXS.pm line 2684.
make[3]: *** [Makefile:424: maxima.info] Error 21
make[3]: Leaving directory '/tmp/portage/sci-mathematics/maxima-5.49.0/work/maxima-5.49.0/doc/info/de'
make[2]: *** [Makefile:629: all-recursive] Error 1
make[2]: Leaving directory '/tmp/portage/sci-mathematics/maxima-5.49.0/work/maxima-5.49.0/doc/info'
make[1]: *** [Makefile:409: all-recursive] Error 1
make[1]: Leaving directory '/tmp/portage/sci-mathematics/maxima-5.49.0/work/maxima-5.49.0/doc'
make: *** [Makefile:467: all-recursive] Error 1
```

The problem is a missing `-I` in `doc/info/de/Makefile.am`. Attached patch fixes the problem for me.



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4726 solve doesn't really solve

[Barton W. <wil...@us...>]

Pasting in a call to `factor` at the top of `solve`, allows Maxima to solve such equations

~~~
(defun solve (*exp *var mult &aux (genvar nil) ($derivsubst nil)
		     (exp (float2rat (mratcheck *exp)))
		     (*myvar *var) ($savefactors t))
  (prog (factors *has*var genpairs $dontfactor temp symbol *g *checkfactors* 
	 varlist expsumsplit)
     (let (($ratfac t))
       (setq exp ($factor (ratdisrep (ratf exp))))) ;;<== added $factor
     ;; Cancel out any simple 
~~~

But this causes some bad bugs; for example,

~~~
******************** rtestint.mac: Problem 243 (line 1451) ********************

Input:
integrate(sqrt(cos(x) + 1), x, - %pi, %pi)


Result:
0

This differed from the expected result:
 5/2
2
~~~


---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Thu Apr 30, 2026 04:05 PM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] Re: #4727 - is converted to +- (in a special case)

[Matthias D. D. <mdi...@us...>]

Maxima itself gives this output, and wxMaxima only casts it in a slightly
different way.
So I think the problem is in Maxima, not in wxMaxima.
I originally had a more complicated expression where I spotted this
behaviour.
I then reduced the expression gradually to a minimal expression which still
gives the weird looking result.
If you write -h instead of -1*h everything is fine, but that's not the
point. You may exchange 1 with any other number.
It seems to me that any further reduction of the example will not show such
weird output.
For example:
1+1-1*h;
gives 2-h (as expected).
But assigning the same expression to a function as
L(h):=1+1-1*h;
gives
L(h) := 1 + 1 + (- 1) h
Of course, I think it is a minor bug, since the result is not wrong.
The question is why
L(h):=1+1-1*h;
is not simplified to
L(h) := 2-h
And why
L(h):=1+1-2*h;
is not simplified to
L(h):=2-2*h

On Fri, 1 May 2026 at 03:30, Stavros Macrakis <
mac...@us...> wrote:

> Strange, it seems that it uses (-1)*h except for the *second* term, where
> it uses -(1*h):
>
> ?print(-1*h)$((MTIMES) ((MMINUS) 1) $H)   ; -(1*h)
>
> ?print(-1*h-1*h)$((MPLUS) ((MTIMES) ((MMINUS) 1) $H) ((MMINUS) ((MTIMES) 1 $H))) ; -(1*h)
>
> ?print(-1*h-1*h-1*h)$((MPLUS) ((MTIMES) ((MMINUS) 1) $H) ((MMINUS) ((MTIMES) 1 $H))  ; -(1*h) ((MTIMES) ((MMINUS) 1) $H))
>
> ?print(-1*h-1*h-1*h-1*h)$((MPLUS) ((MTIMES) ((MMINUS) 1) $H) ((MMINUS) ((MTIMES) 1 $H))  ; -(1*h) ((MTIMES) ((MMINUS) 1) $H) ((MTIMES) ((MMINUS) 1) $H))
>
> ?print(-1*h-1*h-1*h-1*h-1*h);((MPLUS) ((MTIMES) ((MMINUS) 1) $H) ((MMINUS) ((MTIMES) 1 $H))  ; -(1*h) ((MTIMES) ((MMINUS) 1) $H) ((MTIMES) ((MMINUS) 1) $H) ((MTIMES) ((MMINUS) 1) $H))
>
> Surely the implementor didn't intend this, so I guess it's a bug.
> But since they simplify the same, it doesn't really matter.
> ------------------------------
>
> *[bugs:#4727] <https://sourceforge.net/p/maxima/bugs/4727/> - is converted
> to +- (in a special case)*
>
> *Status:* open
> *Group:* None
> *Created:* Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
> *Last Updated:* Thu Apr 30, 2026 06:18 PM UTC
> *Owner:* nobody
>
> *Minimal example*
>
> Entering
>
> L(h):=1+1-1*h;
>
> gives the output
> L(h) := 1 + 1 + (- 1) h
>
> The expected output is
>
> L(h):=1+1-1 h
>
> I spotted the bug in wxMaxima, which displays L(h):=1+1+-1*h
> ------------------------------
>
> Sent from sourceforge.net because you indicated interest in
> https://sourceforge.net/p/maxima/bugs/4727/
>
> To unsubscribe from further messages, please visit
> https://sourceforge.net/auth/subscriptions/
>



---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Thu Apr 30, 2026 07:30 PM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4727 - is converted to +- (in a special case)

[Stavros M. <mac...@us...>]

Strange, it seems that it uses ``(-1)*h`` except for the *second* term, where it uses ``-(1*h)``:
~~~
?print(-1*h)$
((MTIMES) ((MMINUS) 1) $H)   ; -(1*h)

?print(-1*h-1*h)$
((MPLUS)
 ((MTIMES) ((MMINUS) 1) $H)
 ((MMINUS) ((MTIMES) 1 $H))) ; -(1*h)

?print(-1*h-1*h-1*h)$
((MPLUS)
 ((MTIMES) ((MMINUS) 1) $H)
 ((MMINUS) ((MTIMES) 1 $H))  ; -(1*h)
 ((MTIMES) ((MMINUS) 1) $H)) 

?print(-1*h-1*h-1*h-1*h)$
((MPLUS)
 ((MTIMES) ((MMINUS) 1) $H)
 ((MMINUS) ((MTIMES) 1 $H))  ; -(1*h)
 ((MTIMES) ((MMINUS) 1) $H)
 ((MTIMES) ((MMINUS) 1) $H)) 

?print(-1*h-1*h-1*h-1*h-1*h);
((MPLUS)
 ((MTIMES) ((MMINUS) 1) $H)
 ((MMINUS) ((MTIMES) 1 $H))  ; -(1*h)
 ((MTIMES) ((MMINUS) 1) $H)
 ((MTIMES) ((MMINUS) 1) $H)
 ((MTIMES) ((MMINUS) 1) $H))
~~~

Surely the implementor didn't intend this, so I guess it's a bug.
But since they simplify the same, it doesn't really matter.




---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Thu Apr 30, 2026 06:18 PM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4727 - is converted to +- (in a special case)

[Robert D. <rob...@us...>]

A slightly simpler way to write the original example.
```
(%i10) simp: false;
(%o10)                        false
(%i11) 1 + 1 - 1*h;
(%o11)                   1 + 1 + (- 1) h
```
It's interesting to me that was parsed essentially as 1 plus 1 plus ((minus 1) times h):
```
(%i12) :lisp $_

((MPLUS) 1 1 ((MTIMES) ((MMINUS) 1) $H))
```
I wonder why it wasn't parsed as 1 plus 1 plus (minus (1 times h)). I suppose the two forms are equivalent, but a different input does produce what I expected:
```
(%i13) 1 - 1*h;
(%o13)                       1 - 1 h
(%i14) :lisp $_

((MPLUS) 1 ((MMINUS) ((MTIMES) 1 $H)))
```
That is, 1 plus (minus (1 times h)).

I agree that the observed behavior isn't incorrect, I am just slightly mystified about the different ways of parsing a similar expression.


---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Thu Apr 30, 2026 01:26 AM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4726 solve doesn't really solve

[Stavros M. <mac...@us...>]

Thanks for explicitly showing us the settings and versions you're using.

There are a couple of things here which look wrong:

One of the disadvantages of the (...) approach to setting multiple options is that you don't see any errors.
~~~
(assume_pos:true,
 ...
 rootsconmode=super,  << this returns the equation rootsconmode=super,
                              << where you presumably intended the assignment rootsconmode:super
 ...);

(assume(n>2,
 m, constant),        << the returned value of this is
                                                            << [n > 2, meaningless, meaningless]
                                                            << showing that the "m,constant" had no effect
                      << You probably intended a separate declare(m,constant)
~~~
 


---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Thu Apr 30, 2026 11:27 AM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4726 solve doesn't really solve

[Barton W. <wil...@us...>]

Here is a minimal non-working example:

~~~
(%i1)	xxx : 1 = (1-y/2)^(n);
(xxx)	1=(1-y/2)^n

(%i2)	solve(xxx,y), solveexplicit : false;
(%o2)	[(2-y)^n/2^n=1]

(%i3)	solve(%,y);
"Is "n" an "integer"?"n;
(%o3)	[y=0]
~~~

Calling `solve` twice results in one solution (but not all solutions).

When the option variable `solveexplicit`, Maxima realizes that it is unable to solve the equation and it returns the empty list.  Unfortunately,  Maxima's solve function doesn't distinguish the cases of  the solution set is empty and unable to solve:
~~~
(%i1)	xxx : 1 = (1-y/2)^(n);
(xxx)	1=(1-y/2)^n

(%i2)	solve(xxx,y), solveexplicit : true;
(%o2)	[]
~~~

This is a deficiency  of Maxima, not a bug.  Maybe you would like to file a feature request.  If you do, please try to include a simple example that shows what you want. 

For clarity, unless it is necessary to use a non-default value for an option variable, I suggest using the default values--it makes an example easier for me to understand. 



---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Thu Apr 30, 2026 02:39 AM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] Re: #4726 solve doesn't really solve

[Stavros M. <mac...@us...>]

assume(n,integer)
    => [meaningless,meaningless]


On Wed, Apr 29, 2026, 22:39 dan hayes via Maxima-bugs <
max...@li...> wrote:

> i also added assume(n,integer), declare(n,integer) at the beginning and it
> made no difference.
> Also there is the multivalue issue of n+1 different roots but it goes
> without saying the desired one is the principle value or for n even the
> only real root
> ------------------------------
>
> *[bugs:#4726] <https://sourceforge.net/p/maxima/bugs/4726/> solve doesn't
> really solve*
>
> *Status:* open
> *Group:* None
> *Created:* Wed Apr 29, 2026 09:28 AM UTC by dan hayes
> *Last Updated:* Wed Apr 29, 2026 09:28 AM UTC
> *Owner:* nobody
>
> WxMaxima version: 26.01.0_MSW
> Using wxWidgets version: wxWidgets 3.2.9
> Maxima version: 5.49.0
> Maxima build date: 2026-01-02 21:27:51
> Host type: x86_64-w64-mingw32
> System type: Win32 10.0.19041 X86-64
> Lisp implementation type: SBCL
> Lisp implementation version: 2.6.0
>
> (assume_pos:true, file_output_append:true, ratprint:false, showtime:true,
> load(simplify_sum), intanalysis: false, simpsum:true, load("lrats")
> ,letrat:true, ratfac:true,algebraic:true,
> rootsconmode=super,algexact:true,fpprintprec:4);
>
>
> (assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));
>
> Now from the output rs should be in the form y=some function of x but it
> isn't so i had to do
>
> [p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];
>
> to get the correct answer. I tried even using solveradcan:true , which was
> default false and even that
> did not even help to get maxima to give the correct answer. I shouldn't
> actually say "the correct"
> answer because there is an arbitrary constant of integration that could be
> used for expression rx in
> rx:M*integrate(1/(1-x/a),x). Since actually i desire the solution such
> that y=0 when x=0 and a text
> then by some mysterious way to get the answer in the form
>
> ["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];
>
> Now i would not necessarily expect maxima to be able to get this final
> form which is an ungodly
> difficult problem in finding the correct constant of integration to get
> the answer in this form but
> it would be nice. Though if anyone knows how to find that constant of
> integration would be nice. I
> also add that many times prior i have had to do it myself rationalizing
> etc. etc. to get the answer from
> maximas unwillingness to fully solve a straightforward problem. This is
> also a feature request to get
> maxima to fully solve the problem rather than just giving an incomplete
> answer.
> ------------------------------
>
> Sent from sourceforge.net because max...@li... is
> subscribed to https://sourceforge.net/p/maxima/bugs/
>
> To unsubscribe from further messages, a project admin can change settings
> at https://sourceforge.net/p/maxima/admin/bugs/options. Or, if this is a
> mailing list, you can unsubscribe from the mailing list.
> _______________________________________________
> Maxima-bugs mailing list
> Max...@li...
> https://lists.sourceforge.net/lists/listinfo/maxima-bugs
>



---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Thu Apr 30, 2026 02:39 AM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

Re: [Maxima-bugs] [maxima:bugs] #4726 solve doesn't really solve

[Stavros M. <mac...@gm...>]

assume(n,integer)
    => [meaningless,meaningless]


On Wed, Apr 29, 2026, 22:39 dan hayes via Maxima-bugs <
max...@li...> wrote:

> i also added assume(n,integer), declare(n,integer) at the beginning and it
> made no difference.
> Also there is the multivalue issue of n+1 different roots but it goes
> without saying the desired one is the principle value or for n even the
> only real root
> ------------------------------
>
> *[bugs:#4726] <https://sourceforge.net/p/maxima/bugs/4726/> solve doesn't
> really solve*
>
> *Status:* open
> *Group:* None
> *Created:* Wed Apr 29, 2026 09:28 AM UTC by dan hayes
> *Last Updated:* Wed Apr 29, 2026 09:28 AM UTC
> *Owner:* nobody
>
> WxMaxima version: 26.01.0_MSW
> Using wxWidgets version: wxWidgets 3.2.9
> Maxima version: 5.49.0
> Maxima build date: 2026-01-02 21:27:51
> Host type: x86_64-w64-mingw32
> System type: Win32 10.0.19041 X86-64
> Lisp implementation type: SBCL
> Lisp implementation version: 2.6.0
>
> (assume_pos:true, file_output_append:true, ratprint:false, showtime:true,
> load(simplify_sum), intanalysis: false, simpsum:true, load("lrats")
> ,letrat:true, ratfac:true,algebraic:true,
> rootsconmode=super,algexact:true,fpprintprec:4);
>
>
> (assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));
>
> Now from the output rs should be in the form y=some function of x but it
> isn't so i had to do
>
> [p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];
>
> to get the correct answer. I tried even using solveradcan:true , which was
> default false and even that
> did not even help to get maxima to give the correct answer. I shouldn't
> actually say "the correct"
> answer because there is an arbitrary constant of integration that could be
> used for expression rx in
> rx:M*integrate(1/(1-x/a),x). Since actually i desire the solution such
> that y=0 when x=0 and a text
> then by some mysterious way to get the answer in the form
>
> ["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];
>
> Now i would not necessarily expect maxima to be able to get this final
> form which is an ungodly
> difficult problem in finding the correct constant of integration to get
> the answer in this form but
> it would be nice. Though if anyone knows how to find that constant of
> integration would be nice. I
> also add that many times prior i have had to do it myself rationalizing
> etc. etc. to get the answer from
> maximas unwillingness to fully solve a straightforward problem. This is
> also a feature request to get
> maxima to fully solve the problem rather than just giving an incomplete
> answer.
> ------------------------------
>
> Sent from sourceforge.net because max...@li... is
> subscribed to https://sourceforge.net/p/maxima/bugs/
>
> To unsubscribe from further messages, a project admin can change settings
> at https://sourceforge.net/p/maxima/admin/bugs/options. Or, if this is a
> mailing list, you can unsubscribe from the mailing list.
> _______________________________________________
> Maxima-bugs mailing list
> Max...@li...
> https://lists.sourceforge.net/lists/listinfo/maxima-bugs
>

[Maxima-bugs] [maxima:bugs] #4726 solve doesn't really solve

[dan h. <zm...@us...>]

i also added assume(n,integer), declare(n,integer) at the beginning and it made no difference. 
Also there is the multivalue issue of n+1 different roots but it goes without saying  the desired one is the principle value or for n even the only real root


---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Wed Apr 29, 2026 09:28 AM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4727 - is converted to +- (in a special case)

[Stavros M. <mac...@us...>]

The form `(-1)*h` is a faithful rendition of the unsimplified internal form.
It's not pretty, but it's not wrong.
I'm curious, why did you write `-1*h`rather than `-h`? 

What version of Maxima/Lisp/WxMaxima are you running?


---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Thu Apr 30, 2026 01:00 AM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] Re: #4727 - is converted to +- (in a special case)

[Stavros M. <mac...@us...>]

The body of a named or anonymous (lambda) function definition is never simplified.



---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Wed Apr 29, 2026 11:21 PM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4727 - is converted to +- (in a special case)

[Raymond T. <rt...@us...>]

I wonder why `1+1-1*h` isn't simplified to `2-h`.  Nevertheless, the fact that wxmaxima displays it as `1+1+-1*h` seems  to be a wxmaxima bug.


---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Wed Apr 29, 2026 09:23 PM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4727 - is converted to +- (in a special case)

[Matthias D. D. <mdi...@us...>]

---

**[bugs:#4727] - is converted to +-    \(in a special case\)**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:22 PM UTC by Matthias Daniel Diehl
**Last Updated:** Wed Apr 29, 2026 09:22 PM UTC
**Owner:** nobody


**Minimal example**

Entering

L(h):=1+1-1*h;

gives the output
L(h) := 1 + 1 + (- 1) h

The expected output is

L(h):=1+1-1 h

I spotted the bug in wxMaxima, which displays     L(h):=1+1+-1*h



---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.

[Maxima-bugs] [maxima:bugs] #4726 solve doesn't really solve

[dan h. <zm...@us...>]

---

**[bugs:#4726] solve doesn't really solve**

**Status:** open
**Group:** None
**Created:** Wed Apr 29, 2026 09:28 AM UTC by dan hayes
**Last Updated:** Wed Apr 29, 2026 09:28 AM UTC
**Owner:** nobody


WxMaxima version:             26.01.0_MSW
Using wxWidgets version:      wxWidgets 3.2.9
Maxima version:               5.49.0
Maxima build date:            2026-01-02 21:27:51
Host type:                    x86_64-w64-mingw32
System type:                  Win32 10.0.19041 X86-64
Lisp implementation type:     SBCL
Lisp implementation version:  2.6.0

(assume_pos:true, file_output_append:true, ratprint:false,  showtime:true, load(simplify_sum), intanalysis: false, simpsum:true,  load("lrats") ,letrat:true, ratfac:true,algebraic:true, rootsconmode=super,algexact:true,fpprintprec:4);

(assume(n>2,m,constant),ry:integrate((1-y/b)^n,y),rx:M*integrate(1/(1-x/a),x),rs:solve(ry-rx,y));

Now from the output rs should be in the form y=some function of x but it isn't so i had to do

[p12:part(rs,1,2)*b^(n+1),t:p12^(1/(n+1)),ans:b-t,"y"=ans];

to get the correct answer. I tried even using solveradcan:true , which was default false and even that
did not even help to get maxima to give the correct answer. I shouldn't actually say "the correct" 
answer because there is an arbitrary constant of integration that could be used for expression rx in
rx:M*integrate(1/(1-x/a),x). Since actually  i desire the solution such that y=0 when x=0 and a text 
then by some mysterious way to get the answer in the form

["y"=b-b*(1+(n+1)*M*a/b*log(1-x/a))^(1/(n+1))];  

Now i would not necessarily expect maxima to be able to get this final form which is an ungodly 
difficult problem in finding the correct  constant of integration to get the answer in this form but
it would be nice. Though if anyone knows how to find that constant of integration would be nice. I 
also add that many times prior i have had to do it myself rationalizing etc. etc. to get  the answer from
maximas unwillingness to fully solve a  straightforward problem. This is also a feature request to get
maxima to fully solve the problem rather than just giving an incomplete answer.








---

Sent from sourceforge.net because max...@li... is subscribed to https://sourceforge.net/p/maxima/bugs/

To unsubscribe from further messages, a project admin can change settings at https://sourceforge.net/p/maxima/admin/bugs/options.  Or, if this is a mailing list, you can unsubscribe from the mailing list.