Tom Jackson said:
> Yes, catch should not be used unless you need to do something important
I think you are taking me too literal. When an uncaught error occurs (ie:
a situation you never thought would happen), the interpretter returns on
error. This is when AOLserver continues and does what it needs to do
(invoke an error page, log the error). This is where I was trying to hook
in to get some meaningfull reporting. Simply using errorInfo in
errorpage.adp (which seems to be interpretted in the same thread, without
cleaning it up first) works like a charm.

The benefit of doing it this way is that AOLserver only logs the Tcl stack
trace. Having your own handler, you can report not just that, but also the
exact URL and GET data (possibly POST too, but be carefull this could be
big), any headers sent in the request (browser version, referer, cookies)
and other application data you have, like user id. This makes debugging so
much easier.

I remember once asking my brother about some weird errors from OpenACS.
His reply? "oh, don't worry about them, probably some robot sending bad
form data" I am sure it could very well be, but had I had more info, I
could have seen it for myself, seen what was bad about the data and fix
the code so next time it doesn't barf on it and wastes my time.

Bas.


--
AOLserver - http://www.aolserver.com/

To Remove yourself from this list, simply send an email to <[EMAIL PROTECTED]> with the
body of "SIGNOFF AOLSERVER" in the email message. You can leave the Subject: field of 
your email blank.

Reply via email to