Re: [PATCH] Small fixes to AWT

[David Jee <djee at redhat dot com>]

On Tue, 2004-02-10 at 12:10, David Jee wrote:
> On Tue, 2004-02-10 at 11:53, Thomas Fitzsimmons wrote:
> > On Tue, 2004-02-10 at 11:07, David Jee wrote:
> > >         * java/awt/Component.java
> > >         (show): Invalidate after show or hide.
> > > 
> > 
> > What problem does this one fix?  If invalidation really is needed
> > shouldn't it occur in show() and hide() themselves, since they can be
> > called directly?
> 
> When setVisible() is called on a component, we need to invalidate the
> component and its parent container, in order to ensure that the parent
> container's children are laid out properly to reflect the change.  For
> example, in BorderLayout, if you have two visible components, one in
> CENTER and one in EAST, and you setVisible(false) on the EAST component,
> the EAST container should disappear and the CENTER component should span
> the entire width of the parent container.  If you setVisible(true) on
> the EAST component, then the CENTER component should shrink to make room
> for the EAST component.
> 
> setVisible(boolean) is an alias method for the deprecated show(boolean),
> which calls show() or hide() depending on the boolean value.  Regardless
> of the boolean value, we should invalidate the component.

Actually, I finally understand what you mean by invalidating in show or
hide.  D'oh.  I'll make that change.

-David Jee