 |
Matt S Trout |
at Feb 2, 2008 at 6:40 am
|
⇧ | 
| |
On Fri, Feb 01, 2008 at 05:12:12PM -0000, Carl Vincent wrote:http://trout.me.uk/perl/Catalyst-Plugin-StackTrace-0.07.tar.gzespecially if you've seen the bug.
If it all looks good, I'll upload the new version to CPAN (and
if it turns out to be buggy and you -don't- test it now it's
your fault when your application breaks :).
I was having trouble with this yesterday and you helped workaround on
#catalyst. I've tried this new version with my troublesome code of
yesteday and now the exceptions are working with and without
Plugin::StackTrace enabled.
The thing is, the bit that was causing it is stringification, which I'm not
sure is really that useful for debugging.
Should we turn respect_overload off on Stacktrace (or off by default with a
config option if somebody wants to write that code?) ?
I'm actually tempted now to clean the module up a bit now people are
reporting it's fixed, since with my improved understanding of just how fucking
insane SIGDIE handlers are I think I can get rid of a bunch of code that
was supposed to make things safer but aren't strictly required :)
