[Catalyst] [ANNOUNCE] Catalyst::Plugin::StackTrace

Brandon Black blblack at gmail.com
Sat Dec 31 19:02:26 CET 2005


On 12/31/05, Matt S Trout <dbix-class at trout.me.uk> wrote:
> On Sat, Dec 31, 2005 at 04:12:29PM +0100, Johan Lindstr�m wrote:
> > At 05:18 2005-12-31, Andy Grundman wrote:
> > >I've pushed the first version of the StackTrace plugin to CPAN.
> > >
> > >Screenshot: http://www.hybridized.org/catalyst/stacktrace.png
> > >
> > >Download:
> > >http://pause.perl.org/incoming/Catalyst-Plugin-StackTrace-0.01.tar.gz
> >
> > Very, very nice!
> >
> > However, I had to foce install it because of a breaking test (the test
> > seems broken, not the module). I submitted an RT ticket on CPAN with the
> > details.
> >
> > Ok, I haven't used it for that many minutes yet, but IMHO this
> > functionality is so useful and so central to Catalyst that it should be
> > promoted to core. Maybe when it has accumulated a few system-weeks of
> > deployment?
>
> Maybe in the sense that Static::Simple is promoted to core - but not
> embedded in the main code. Overriding the __DIE__ and __WARN__ sig handlers
> is always potentially tricky, and if somebody's using an insane exception
> system that needs to do so itself (or that doesn't like the interruption
> to the exception process that StackTrace causes) it has to be trivial to
> turn it off.
>
> Plus StackTrace causes an additional overhead on every thrown exception, which
> in exception-heavy code may be an additional performance hit.
>

I agree with mst on both points.  Nice work Andy :)


More information about the Catalyst mailing list