Simultaneous multiple module versions

[Jonathan Stowe]

On Thu, 2006-08-17 at 16:53, Nicholas Clark wrote:
> On Thu, Aug 17, 2006 at 08:22:48AM -0700, Dan Sully wrote:
> > * Nicholas Clark shaped the electrons to say...
> > 
> > >You've measured this? Or you just read this somewhere?
> > >If the former, do you have a demonstration case?
> > 
> > This "fix" impacted SlimServer a lot. One of my guys ran some tests - we use
> > DBIx::Class underneath.
> 
> It's not a "fix", as it fixes a real bug, and not a Class::DBI specific bug.
> 
> I'm surprised that the speed hit is that bad - this may mean that there's a
> bug in the implementation. But the only change in behaviour *should* be
> when blessing an object from a class without to a class with overloading
> (or from something unblessed into a class with overloading), if and only
> if there is more than one reference to the object.
> 
> So, bless ({}, $class) shouldn't be affected, as there will be only one
> reference to that {}
> 
> No-one's reported the performance hit upstream, let alone with a demonstration
> case that uses only core modules. Is anyone surprised that there's been
> no response?
> 

A possible explanation is that the problem only manifests itself when
the change is applied to an older perl.

> london.pm list isn't the right place either.
> 

Quite.
> Nicholas Clark
-- 

This e-mail is sponsored by http://www.integration-house.com/