Re: Subclassing question

[muppet <scott asofyet org>]

On May 16, 2009, at 12:42 PM, Torsten Schoenfeld wrote:

> Technically, this is due to gperl_value_from_sv using  
> g_value_set_boxed instead of g_value_set_static_boxed.  The former  
> copies the boxed object.  I can't think of a reason not to use  
> g_value_set_static_boxed there (other than "being defensive").  When  
> I do this change, Glib's and Gtk2's test suites still pass.
>
> muppet, do you remember why gperl_value_from_sv uses  
> g_value_set_boxed instead of g_value_set_static_boxed?

Initially, we assumed that a boxed object always holds a copy.  I was  
blissfully unaware of various deep subtleties when that code was  
new.  :-)  That gets wrapped up in "are you allowed to change this  
thing through your reference?" questions, but i think the important  
ones of those are taken care of for us upstream.

If that actually works, then i think we get a functionality win (some  
of the marshaler behave more usefully) that would be worth some bug  
reports, so i think it's worth pushing it out in an unstable release  
to see what happens.

Specific things to test would include custom treemodels and complex  
perl widget or cellrenderer implementations.


--
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"