#5684 Change in Maxima versions

[Allan James Payne]

The attached file works successfully to completion on 16Sep26 but fails on 15Oct26 on
WxMaxima version: 26.08.0_MSW
Using wxWidgets version: wxWidgets 3.3.3
Maxima version: branch_5_50_base_9_gf03405fbf_dirty
Maxima build date: 2026-08-18 22:11:52
Host type: x86_64-w64-mingw32
System type: Win32 10.0.19045 X86-64
Lisp implementation type: SBCL
Lisp implementation version: 2.6.7
with message
region_profile_3342["rISCO"]
-- an error. To debug this try: debugmode(true);
on line
rp : region_manager([M, a, Q], -1, 50, ["tol" : 1e-8])$

[Allan James Payne]

Hi David
Thanks for your email. Claude is great for diagnosis. I have used Copilot
and the google freeby.
Thanks for the diagnosis. I will take the project forward now. I attach my
previous working version for your interest.
Rgds Allan Payne

On Thu, Oct 15, 2026 at 12:05 PM David Scherfgen tomasriker@users.sourceforge.net wrote:

• status: open --> not-a-bug
• Comment:

By Claude:

This isn't a Maxima bug. The line rp : region_manager([M, a, Q], -1, 50,
["tol" : 1e-8])$ fails because of errors in the script itself:

1. ["tol" : 1e-8] is a list containing an assignment, and the
   assignment runs when the list is built. Assigning to a string is an error (assignment:
   cannot assign to string "tol"). Use an equation instead: ["tol"
   = 1e-8]. The same goes for meta : ["spin_sign" : sgn,
   "solver_opts" : opts] in build_region_kn. In build_region_kerr,
   [spin_sign : sgn] doesn't fail, but it sets a global variable spin_sign
   and the list is just [1].
2. Even with that fixed, the call can't succeed. With three
   parameters, region_manager ends up in rgamma_ret_kn, a placeholder
   that deliberately calls error("rgamma_ret_kn: not yet implemented
   ..."). Remove that line or wrap it in errcatch(...) until the
   Kerr–Newman solvers are implemented.

The message is confusing because errormsg:false (right after
load("ctensor")) suppresses the actual error message.
region_profile_3342["rISCO"] is just the output of the preceding
print(rp_rISCO(rp)). Remove errormsg:false to see the real messages.

Side note: the second set of accessors (rp_rISCO(rp) :=
rp["rISCO"] etc.) replaces the get()-based ones defined just
before it. rp["rISCO"] subscripts the symbol instead of reading
its properties, which is why it prints
region_profile_3342["rISCO"] instead of the radius. Deleting
those definitions fixes that.

I ran the file with Maxima built from the same commit as yours (f03405fbf)
and with current master. Both stop at that line in the same way. With the
line wrapped in errcatch, the rest of the file runs to the end on both,
with identical results. Since the script itself calls error() there, that
line can't have succeeded with any Maxima version. So something in the
script, or in the way it was run, must have changed since 16 Sep.

---

[bugs:#5684] https://sourceforge.net/p/maxima/bugs/5684/ Change in
Maxima versions

Status: not-a-bug
Group: None
Created: Thu Oct 15, 2026 10:10 AM UTC by Allan James Payne
Last Updated: Thu Oct 15, 2026 10:11 AM UTC
Owner: nobody
Attachments:

• blackholetest.txt
  https://sourceforge.net/p/maxima/bugs/5684/attachment/blackholetest.txt
  (sourceforge.net) (51.2 kB; text/plain)

The attached file works successfully to completion on 16Sep26 but fails on
15Oct26 on
WxMaxima version: 26.08.0_MSW
Using wxWidgets version: wxWidgets 3.3.3
Maxima version: branch_5_50_base_9_gf03405fbf_dirty
Maxima build date: 2026-08-18 22:11:52
Host type: x86_64-w64-mingw32
System type: Win32 10.0.19045 X86-64
Lisp implementation type: SBCL
Lisp implementation version: 2.6.7
with message
region_profile_3342["rISCO"]
-- an error. To debug this try: debugmode(true);
on line
rp : region_manager([M, a, Q], -1, 50, ["tol" : 1e-8])$

---

Sent from sourceforge.net because you indicated interest in
https://sourceforge.net/p/maxima/bugs/5684/

To unsubscribe from further messages, please visit
https://sourceforge.net/auth/subscriptions/

[David Scherfgen]

By Claude:

This isn't a Maxima bug. The line rp : region_manager([M, a, Q], -1, 50, ["tol" : 1e-8])$ fails because of errors in the script itself:

1. ["tol" : 1e-8] is a list containing an assignment, and the assignment runs when the list is built. Assigning to a string is an error (assignment: cannot assign to string "tol"). Use an equation instead: ["tol" = 1e-8]. The same goes for meta : ["spin_sign" : sgn, "solver_opts" : opts] in build_region_kn. In build_region_kerr, [spin_sign : sgn] doesn't fail, but it sets a global variable spin_sign and the list is just [1].
2. Even with that fixed, the call can't succeed. With three parameters, region_manager ends up in rgamma_ret_kn, a placeholder that deliberately calls error("rgamma_ret_kn: not yet implemented ..."). Remove that line or wrap it in errcatch(...) until the Kerr–Newman solvers are implemented.

The message is confusing because errormsg:false (right after load("ctensor")) suppresses the actual error message. region_profile_3342["rISCO"] is just the output of the preceding print(rp_rISCO(rp)). Remove errormsg:false to see the real messages.

Side note: the second set of accessors (rp_rISCO(rp) := rp["rISCO"] etc.) replaces the get()-based ones defined just before it. rp["rISCO"] subscripts the symbol instead of reading its properties, which is why it prints region_profile_3342["rISCO"] instead of the radius. Deleting those definitions fixes that.

I ran the file with Maxima built from the same commit as yours (f03405fbf) and with current master. Both stop at that line in the same way. With the line wrapped in errcatch, the rest of the file runs to the end on both, with identical results. Since the script itself calls error() there, that line can't have succeeded with any Maxima version. So something in the script, or in the way it was run, must have changed since 16 Sep.