[openbeos] Re: BBox bug

that's what should happen, but if the app server cleared the background then 
bevel would be cleared as well, unless it clips against the update rect & 
there's something wrong with calculating the rect to be updated...

Mahmoud

>Selon Mahmoud Al Gammal <obos@xxxxxxxxxxx>:

>> There's a problem with our implementation of BBox (in app_kit/source/lib/
>> interface/src). When the box is set to B_FOLLOW_ALL_SIDES and the window 
is 
>> resized the old border leaves traces all over the box's background.
>> I've taken a look at the code & I think there's nothing wrong with it. 
>> Actually, I only tried it because I had the very same problem with a BView 
>> subclass that I'm implementing, so I thought I'd find a solution in our 
>BBox,
>> 
>> but now it needs to be fixed too!
>> 
>hmm, TAB is dangerous...

>so, 
>actually, R5 has a similar bug I think (or was it R5.1 ?), 
>I saw it on an app where a toolbar was using a BBox as 
>background view.
>A possible fix would be to Invalidate() all the view on resize,
>but that wouldn't be very nice to look at... another possibility 
>is to Invalidate() / Draw() and Flush() in the FrameResize() 
>hook, if only the old size is still available.

>Btw, this happens because, when resizing, app_server doesn't 
>force a redraw of the whole frame, neither does BBox itself,
>it just redraws the bevel, after app_server has painted the 
>ViewColor on the newly exposed area.

>François.

Other related posts: