Re: chaining to superclass in CUSTOM_TAG_START

[Kevin Ryde <user42 zip com au>]

muppet <scott asofyet org> writes:
> SUPER::METHOD_NAME in perl...

I see getting the right start point for a search upwards is a bit of a
pain.  The C funcs like gtk_container_buildable_custom_tag_start() know
where they belong, but say gtk2perl_buildable_custom_tag_start() is
shared.

> forgive me if i punt

"Punt" meaning "kick it into touch" :-) ... there's not really a direct
equivalent in aussie rules now with the deliberate-out-of-bounds rule.

Perhaps a caution in the docs though about what's lost.  I think (still)
chaining up would be the most helpful default, if it could be done.

Index: GtkBuildable.xs
===================================================================
--- GtkBuildable.xs     (revision 2080)
+++ GtkBuildable.xs     (working copy)
@@ -868,6 +868,14 @@
 interface to your object declaration, and implement the following methods
 as necessary.
 
+Note that in the current implementation the custom tags code doesn't
+chain up to any buildable interfaces in superclasses.  This means for
+instance if you implement Gtk2::Buildable on a new widget subclass
+then you lose the <accelerator> and <accessibility> tags normally
+available from Gtk2::Widget.  This will likely change in the future,
+probably by chaining up by default for unhandled tags, maybe with a
+way to ask deliberately not to chain.
+
 =over
 
 =item SET_NAME ($self, $name)


-- 
There are only two classes of people in New South Wales: those who have
been convicted and those who ought to have been.
    -- Governor Macquarie to Lord Bathurst, 1822