On Wed, Jul 11, 2007 at 10:31:22AM -0300, Otávio Fernandes
wrote:
[...]
> If you're using a web application, I recommend
CGI::Carp, because you can
> write an personalized routine to handle application
errors. I my case,
> when my scripts return an error (like an die, or any
other fail), instead
> my users view "Internal Server Error" (from
Apache), they see my error
> routine build with Carp and CGI ... It's very
interesting.
I like using CGI::Carp even for non-CGI programs, because
you get the
timestamp and name of the script which can be awfully handy
if your program
has ended up being launched from cron and buried five levels
deep in a maze
of twisty shell scripts, all alike.
It doesn't half confuse the sysadmins when your Perl daemon
process logs to
syslog instead of STDERR. It's almost as if they've never
used any other
daemonised program in Unix before, or were expecting all
Perl to be
something along the level of Matt Wright's code.
|