[slime-devel] problem with swank-debugger-hook
Hi, I'd like to use slime to debug araneida requests. I did (setf *restart-on-handler-errors* 'swank::debug-in-emacs) But when I hit an error, emacs does not jump in the debugger, and I get the following message on the standard output instead: debugger invoked on a TYPE-ERROR: The value NIL is not of type SWANK::CONNECTION. ...in (SWANK::FORCE-USER-OUTPUT) So, somehow, *EMACS-CONNECTION* is NIL at the time of the call. Don't know why however... Has anyone a clue or a workaround ? This is with the latest CVS. Frédéric
frederic.gobry@epfl.ch (Frédéric Gobry) writes:
Hi,
I'd like to use slime to debug araneida requests. I did
(setf *restart-on-handler-errors* 'swank::debug-in-emacs)
But when I hit an error, emacs does not jump in the debugger, and I get the following message on the standard output instead:
try using #'swank:swank-debugger-hook instead. -- -Marco Ring the bells that still can ring. Forget the perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen
try using #'swank:swank-debugger-hook instead.
Is it ok to pass nil as the "hook" argument? So far, the hook seems not to be used, but this does not tell me what I should put there to be safe :) (setf araneida:*restart-on-handler-errors* #'(lambda (condition) (swank:swank-debugger-hook condition nil))) Frédéric
frederic.gobry@epfl.ch (Frédéric Gobry) writes:
try using #'swank:swank-debugger-hook instead.
Is it ok to pass nil as the "hook" argument? So far, the hook seems not to be used, but this does not tell me what I should put there to be safe :)
the hook argument to swank-debugger-hook is ignored, so pass whatever you want. why don't we just remove the argument if it's ignored? are we sure it sholud be ignored? -- -Marco Ring the bells that still can ring. Forget the perfect offering. There is a crack in everything. That's how the light gets in. -Leonard Cohen
On 9217 day of my life Marco Baringer wrote:
why don't we just remove the argument if it's ignored? are we sure it sholud be ignored?
Make it optional. -- Ivan Boldyrev Outlook has performed an illegal operation and will be shut down. If the problem persists, contact the program vendor.
"Marco Baringer" <mb@bese.it> writes:
frederic.gobry@epfl.ch (Frédéric Gobry) writes:
try using #'swank:swank-debugger-hook instead.
Is it ok to pass nil as the "hook" argument? So far, the hook seems not to be used, but this does not tell me what I should put there to be safe :)
the hook argument to swank-debugger-hook is ignored, so pass whatever you want.
why don't we just remove the argument if it's ignored? are we sure it sholud be ignored?
We can't remove it because CL:*DEBUGGER-HOOK* needs to be bound to a function that takes two arguments.
"Marco Baringer" <mb@bese.it> writes:
frederic.gobry@epfl.ch (Frédéric Gobry) writes:
try using #'swank:swank-debugger-hook instead.
Is it ok to pass nil as the "hook" argument? So far, the hook seems not to be used, but this does not tell me what I should put there to be safe :)
the hook argument to swank-debugger-hook is ignored, so pass whatever you want.
why don't we just remove the argument if it's ignored? are we sure it sholud be ignored?
If we're offering the possibility of using swank-debugger-hook as CL:*DEBUGGER-HOOK*, then it needs to be a function of two arguments. *DEBUGGER-HOOK* is bound to nil when calling the current debugger hook; it may be necessary to introduce (let ((*debugger-hook* hook)) ...) around debugger hook activities that may invoke the debugger... Cheers, Christophe
participants (5)
-
Christophe Rhodes -
frederic.gobry＠epfl.ch -
Ivan Boldyrev -
Luke Gorrie -
Marco Baringer