stack-trace-on-error vs. debug-on-error.

stack-trace-on-error vs. debug-on-error.

[Lute Kamstra]

Does anyone know the difference between stack-trace-on-error and
debug-on-error?

,----[ C-h v stack-trace-on-error RET ]
| stack-trace-on-error's value is nil
| 
| *Non-nil means errors display a backtrace buffer.
| More precisely, this happens for any error that is handled
| by the editor command loop.
| If the value is a list, an error only means to display a backtrace
| if one of its condition symbols appears in the list.
| 
| You can customize this variable.
| 
| Defined in `C source code'.
`----

,----[ C-h v debug-on-error RET ]
| debug-on-error's value is nil
| 
| *Non-nil means enter debugger if an error is signaled.
| Does not apply to errors handled by `condition-case' or those
| matched by `debug-ignored-errors'.
| If the value is a list, an error only means to enter the debugger
| if one of its condition symbols appears in the list.
| When you evaluate an expression interactively, this variable
| is temporarily non-nil if `eval-expression-debug-on-error' is non-nil.
| See also variable `debug-on-quit'.
| 
| You can customize this variable.
| 
| Defined in `C source code'.
`----

On first inspection they seem to do exactly the same.
stack-trace-on-error is not documented in Emacs manual or the Lisp
manual.  Would it be safe to make it an alias of debug-on-error and
mark it as obsolete?

Lute.

Re: stack-trace-on-error vs. debug-on-error.

[Lute Kamstra]

Lute Kamstra <Lute.Kamstra.lists@xs4all.nl> writes:

> Does anyone know the difference between stack-trace-on-error and
> debug-on-error?

[...]

> On first inspection they seem to do exactly the same.

Silly me: I was testing it in a *scratch* buffer with
eval-expression-debug-on-error on, so errors always took me to the
debugger.  I now see the difference: stack-trace-on-error just prints
a backtrace, while debug-on-error enters the debugger.

> stack-trace-on-error is not documented in Emacs manual or the Lisp
> manual.

I'll document it in the Lisp manual.

Lute.

Re: stack-trace-on-error vs. debug-on-error.

[Lute Kamstra]

Lute Kamstra <Lute.Kamstra.lists@xs4all.nl> writes:

> I'll document it in the Lisp manual.

Here it is:

*** lispref/debugging.texi      3 Mar 2005 16:28:32 -0000       1.27
--- lispref/debugging.texi      4 Mar 2005 11:02:19 -0000
***************
*** 166,171 ****
--- 166,185 ----
            (lambda () (setq debug-on-error t)))
  @end example
  
+ When the debugger is entered, it shows a backtrace (@pxref{Using
+ Debugger}).  If you like to see the backtrace when an error happens,
+ but you do not want to enter the debugger, you can set the variable
+ @code{stack-trace-on-error} to non-@code{nil}.
+ 
+ @defopt stack-trace-on-error
+ This variable determines whether a backtrace buffer is shown when an
+ error is signalled and not handled.  If @code{stack-trace-on-error} is
+ @code{t}, all kinds of errors display a backtrace; if it is
+ @code{nil}, none do.  If the value is a list, an error only means to
+ display a backtrace if one of its condition symbols appears in the
+ list.
+ @end defopt
+ 
  @node Infinite Loops
  @subsection Debugging Infinite Loops
  @cindex infinite loops

Re: stack-trace-on-error vs. debug-on-error.

[drkm]

> Lute Kamstra wrote:

> drkm <darkman_spam@yahoo.fr> writes:

> >     M-: (setq debug-on-error       nil
> >               stack-trace-on-error t)
> >     C-f
> >     ;; Open a new window and show the stack trace

> >     M-: (setq debug-on-error       t
> >               stack-trace-on-error nil)
> >     C-f
> >     ;; Print "End of buffer" in the echo area

> Yes, and when you do

>     M-: (setq debug-on-error       t
>               debug-ignored-errors nil)
>               stack-trace-on-error nil)
>     C-f

> you enter the debugger.

  Maybe displaying the stack trace would care about
debug-ignored-errors, too?

--drkm



	

	
		
Découvrez le nouveau Yahoo! Mail : 250 Mo d'espace de stockage pour vos mails ! 
Créez votre Yahoo! Mail sur http://fr.mail.yahoo.com/

Re: stack-trace-on-error vs. debug-on-error.

[Richard Stallman]

    > stack-trace-on-error is not documented in Emacs manual or the Lisp
    > manual.

    I'll document it in the Lisp manual.

Please don't.  It is not worth taking up space in the manual.
Perhaps it should be marked as obsolete, and perhaps it should
be deleted, but I am not sure.

Re: stack-trace-on-error vs. debug-on-error.

[Chris Smith]

         I'll document it in the Lisp manual.

     Please don't.  It is not worth taking up space in the manual.
     Perhaps it should be marked as obsolete, and perhaps it should
     be deleted, but I am not sure.

That could be an interesting addition to http://www.emacswiki.org
Why not post it there?
Best,
Chris