Re: buildable interface testers error

[muppet <scott asofyet org>]

On Nov 1, 2008, at 2:21 PM, Torsten Schoenfeld wrote:

> muppet wrote:
> > The obvious scenario that comes to mind is a hosed registration  
> > order.  Is there any value in worrying about deferred  
> > initialization?  I don't think so, as you're typically doing  
> > instantiation after initialization has finished, but we have ways  
> > of breaking that assumption.
>
> I don't quite understand what you mean here.  In  
> gperl_register_object we force interfaces to finish loading  
> immediately.  So right after 'use Gtk2', all interfaces should be  
> registered properly.  Oh, do you mean this kind of setup?

[load Gtk2 after trying to use it]

> Yeah, that triggers the warning.  But I think that's fair.

That's the gist, yes.  There are interesting scenarios, usually  
involving eval and other ways to defer execution of "use", that create  
hard-to-debug problems.  In most cases, this wizardry is a bad idea.   
All i'm really interested in is making sure we provide a clue.


--
The front of your fridge may be cheap and nasty, but I hope the next  
time you look at the back of your fridge, you'll regard it with  
suitable admiration.
  -- Tim Hunkin, "The Secret Life of Machines"