No BACKTRACE anymore?

[petter]

I'm running
(asdf:component-version (asdf:find-system :weblocks))
"0.8.3"
(ql:system-apropos "weblocks")
#<SYSTEM weblocks / weblocks-20120811-git / quicklisp 2012-08-11>

And I've started my app using the debug: t flag. It appears to be
enabled in my app (I'm only running a single app):

* (weblocks-webapp-debug (first *active-webapps*))
T

However, whenever I get an error I only get the following output:

Internal Server Error

An error occured while processing your /regrun request.

__________________________________________________________________

Error Message

ERROR

Backtrace

BACKTRACE

__________________________________________________________________

Hunchentoot 1.2.3 running on SBCL 1.0.57

__________________________________________________________________

[made-with-lisp-logo.jpg]

I get no stacktrace. On my console I see nothing but the usual access
log:

127.0.0.1 - [2012-09-02 13:42:17] "GET ...

So how can I enable some debugging? Is there a new stream or logfile
in the later revisions of weblocks which have replaced the debug
feature as described in theTroubleshooting section of
http://trac.common-lisp.net/cl-weblocks/wiki/UserManual?

[Scott L. Burson]

On Sun, Sep 2, 2012 at 5:02 AM, petter <petter...@gmail.com> wrote:

[...] I get no stacktrace. On my console I see nothing but the usual access

log:

127.0.0.1 - [2012-09-02 13:42:17] "GET ...

So how can I enable some debugging?

Something changed in Hunchentoot 1.2 that broke the backtrace page feature.  I don't think anyone has figured out how to get it back.

Personally, I like to do

  (setq hunchentoot:*catch-errors-p* nil)

This causes server threads to enter the SLIME debugger when they get errors, whereupon I can debug them interactively.  Not everyone likes this, though, I guess -- you can wind up with a lot of threads in the debugger.

-- Scott
 

[petter]

Thank you for your tip. I'll check it out.

[petter]

Some extra threads sure beats blindfolded debugging. Thanks!