Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

[Alan Mackenzie]

Hello, Emacs.

Start RC1 with emacs -Q.

M-: debug-on-error returns t.
C-h v debug-on-error states that the variable has value nil.  How can
  this be?

M-: (setq debug-on-error nil) ; in buffer *scratch*, still.
M-: debug-on-error returns t.  How can this be?

Why does this variable behave so strangely?  Or am I going mad?  Or,
even both?  Should I submit a bug report for this?

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

[Robert Weiner]

[-- Attachment #1: Type: text/plain, Size: 326 bytes --]

On Tue, Jul 26, 2016 at 4:34 PM, Alan Mackenzie <acm@muc.de> wrote:

>
> M-: (setq debug-on-error nil) ; in buffer *scratch*, still.
> M-: debug-on-error returns t.  How can this be?
>

​Good catch.  This problem has been there since at least 25.0.94 since I am
running that and it exhibits it as well.

Bob
​

[-- Attachment #2: Type: text/html, Size: 1221 bytes --]

Re: Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

[Dale Snell]

On Tue, 26 Jul 2016 20:34:50 +0000, in message
20160726203450.GA10926@acm.fritz.box, Alan Mackenzie wrote:

> 
> Why does this variable behave so strangely?  Or am I going mad?  Or,
> even both?  Should I submit a bug report for this?
> 

Hello Alan,

Whatever is causing this, it's not new with RC1.  I get the same
results with 24.5.1 (x86_64-redhat-linux-gnu, GTK+ Version 3.18.9)

--Dale

-- 
“Insanity is part of the times!  You must embrace the madness, let
it fire you.”  --Londo Mollari, Babylon 5

Re: Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

[Sven Joachim]

On 2016-07-26 20:34 +0000, Alan Mackenzie wrote:

> Hello, Emacs.
>
> Start RC1 with emacs -Q.
>
> M-: debug-on-error returns t.
> C-h v debug-on-error states that the variable has value nil.  How can
>   this be?
>
> M-: (setq debug-on-error nil) ; in buffer *scratch*, still.
> M-: debug-on-error returns t.  How can this be?

See the documentation for debug-on-error:

,----
| debug-on-error is a variable defined in ‘C source code’.
| Its value is nil
| 
| Documentation:
| [...]
| When you evaluate an expression interactively, this variable
| is temporarily non-nil if ‘eval-expression-debug-on-error’ is non-nil.
`----

> Why does this variable behave so strangely?  Or am I going mad?  Or,
> even both?  Should I submit a bug report for this?

Everything works as designed, AFAICS (by default,
eval-expression-debug-on-error is t).

Cheers,
       Sven

Re: Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

[Robert Weiner]

[-- Attachment #1: Type: text/plain, Size: 301 bytes --]

On Tue, Jul 26, 2016 at 5:08 PM, Sven Joachim <svenjoac@gmx.de> wrote:

> Everything works as designed, AFAICS (by default,
> eval-expression-debug-on-error is t).
>

​So how can one get a look at the top-level value of debug-on-error when
eval-expression-debug-on-error is set to t?

Bob

[-- Attachment #2: Type: text/html, Size: 934 bytes --]

Re: Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

[John Mastro]

Robert Weiner <rsw@gnu.org> wrote:
> On Tue, Jul 26, 2016 at 5:08 PM, Sven Joachim <svenjoac@gmx.de> wrote:
>>
>> Everything works as designed, AFAICS (by default,
>> eval-expression-debug-on-error is t).
>
>
> So how can one get a look at the top-level value of debug-on-error when
> eval-expression-debug-on-error is set to t?

C-h v debug-on-error RET should work (and that's why its
value was different in Alan's experiment).

       John

Re: Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

[Alan Mackenzie]

Hello, Sven.

On Tue, Jul 26, 2016 at 11:08:17PM +0200, Sven Joachim wrote:
> On 2016-07-26 20:34 +0000, Alan Mackenzie wrote:

> > Hello, Emacs.

> > Start RC1 with emacs -Q.

> > M-: debug-on-error returns t.
> > C-h v debug-on-error states that the variable has value nil.  How can
> >   this be?

> > M-: (setq debug-on-error nil) ; in buffer *scratch*, still.
> > M-: debug-on-error returns t.  How can this be?

> See the documentation for debug-on-error:

> ,----
> | debug-on-error is a variable defined in ‘C source code’.
> | Its value is nil
> | 
> | Documentation:
> | [...]
> | When you evaluate an expression interactively, this variable
> | is temporarily non-nil if ‘eval-expression-debug-on-error’ is non-nil.
> `----

> > Why does this variable behave so strangely?  Or am I going mad?  Or,
> > even both?  Should I submit a bug report for this?

> Everything works as designed, AFAICS (by default,
> eval-expression-debug-on-error is t).

Ah.  I see this now.  Thanks for clearing it up!

> Cheers,
>        Sven

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: Emacs 25.1 RC1: incipient insanity, or can't set debug-on-error.

[Stephen Berman]

On Tue, 26 Jul 2016 20:34:50 +0000 Alan Mackenzie <acm@muc.de> wrote:

> Hello, Emacs.
>
> Start RC1 with emacs -Q.
>
> M-: debug-on-error returns t.
> C-h v debug-on-error states that the variable has value nil.  How can
>   this be?
>
> M-: (setq debug-on-error nil) ; in buffer *scratch*, still.
> M-: debug-on-error returns t.  How can this be?
>
> Why does this variable behave so strangely?  Or am I going mad?  Or,
> even both?  Should I submit a bug report for this?

It's in the definition of eval-expression:

    (let ((old-value (make-symbol "t")) new-value)
      ;; Bind debug-on-error to something unique so that we can
      ;; detect when evalled code changes it.
      (let ((debug-on-error old-value))
	(push (eval (macroexpand-all exp) lexical-binding) values)
	(setq new-value debug-on-error))
      ;; If evalled code has changed the value of debug-on-error,
      ;; propagate that change to the global binding.
      (unless (eq old-value new-value)
	(setq debug-on-error new-value))))

eval-last-sexp does essentially the same thing.

Steve Berman