On Tue, 2006-02-14 at 11:02 +0100, Mark Wielaard wrote:
> Hi,
> 
> A long, long, time ago we had the following conversation:
> 
> On Mon, 2005-09-05 at 15:50 -0400, Thomas Fitzsimmons wrote:
> > On Sat, 2005-09-03 at 21:56 +0200, Mark Wielaard wrote:
> > > Is there a reason why the GtkToolkit createImage() methods only return
> > > an ErrorImage on failure in the case of Graphics2D/BufferedImage
> > > availability? This patch makes them behave similar:
> > [...]
> > Can you file a bug for the problem you're trying to fix?
> > 
> > I want to get rid of GtkErrorImage; rather it should just be a GtkImage
> > with an error flag set.
> 
> Riccardo was so good to help me test some alternative patches. And this
> is what we finally came up with:
> 
> 2006-03-14  Mark Wielaard  <[EMAIL PROTECTED]>
> 
>     Fixes bug #23931
>     * gnu/java/awt/peer/gtk/GtkImage.java (errorImage): New static
>     field.
>     (getErrorImage): New static method.
>     * gnu/java/awt/peer/gtk/GtkToolkit.java (GtkErrorImage): Removed.
>     (bufferedImageOrError): Renamed to ...
>     (imageOrError): Renamed from bufferedImageOrError, takes Image.
>     Returns GtkImage.getErrorImage() when argument null.
>     (createImage(String)): Always use imageOrError.
>     (createImage(URL)): Likewise.
>     (createImage(ImageProducer)): Likewise.
>     (createImage(byte[],int,int)): Likewise. 
> 
> See also the discussion in the bug report:
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23931
> 
> OK to commit?

This is great.  Thanks for working through this.  Please commit.

Tom

> 
> Cheers,
> 
> Mark


Reply via email to