GNU bug report logs - #15101
24.3.50; debugger-eval-expression broken

Previous Next

Full log

Message #11 received at 15101 <at> debbugs.gnu.org (full text, mbox):

From: Drew Adams <drew.adams <at> oracle.com>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>, Helmut Eller
 <eller.helmut <at> gmail.com>
Cc: 15101 <at> debbugs.gnu.org
Subject: RE: bug#15101: 24.3.50; debugger-eval-expression broken
Date: Thu, 15 Aug 2013 08:06:37 -0700 (PDT)

> > Start Emacs with: emacs -Q --eval '(let ((foo 123)) (debug))'
> > Then press e. Enter foo to evaluate the local variable foo. But it
> > doesn't work; it only generates the error:
> > debugger-eval-expression: Symbol's value as variable is void: foo
> > This used to work fine in previous versions.
> 
> Indeed, this is a change that will trip up users.  Here's what's
> happening: `e' will now run the code in the context in which the "code
> on the current line" was run.
> 
> This refinement can be useful for dynamically bound variables, but was
> mostly added for lexically bound variables, where it's indispensable.
> 
> So the above recipe works again if you use C-p before `e' so that point
> is now on the top-most line, which stands for "in the context that
> called `debug'".
> 
> I think a good fix is to change debug.el so that point starts on the
> first line of the *Debugger* buffer rather than on the second.

In your judgment, can both the old `e' behavior and the new `e' behavior
be useful, at least for dynamically bound variables?

If so, perhaps consider having two different keys, for the two different
behaviors.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.