[Date Prev][Date Next][Thread Prev][Thread Next][Thread Index]

RE: [XaraXtreme-dev] Strange rendering problem

Yes, I just tried this in the latest release build and I see the problem
just as you describe. Perhaps this is a problem in Gavin's CDraw changes
that were checked in Monday.

Are you using CDraw 4.1 or 4.2 Martin (XaraLX -v)?


> -----Original Message-----
> From: owner-dev@xxxxxxxxxxxxxxxx [mailto:owner-dev@xxxxxxxxxxxxxxxx]
> Behalf Of Martin Wuerthner
> Sent: 29 March 2006 11:23
> To: dev@xxxxxxxxxxxxxx
> Subject: [XaraXtreme-dev] Strange rendering problem
> Is it just me or do others see this rendering problem, too? This is on
> x86, SuSE 10.0, wxWidgets 2.6.3-rc2, source revision 731 without any
> changes, debug build.
> At certain zoom percentages a spurious black background appears that
> is of the size of the bounding box of all objects on the page. The
> background can be wiped away by moving an object over it but reappears
> when the area is scrolled into view again.
> This is how I can provoke the problem reliably:
> 1. Start XaraLX
> 2. Create a new file - it will be displayed at 100% zoom
> 3. Draw an ellipse in the middle of the page and set its size to 500
>    by 300 pixels.
> 4. Give it a red fill (not necessary to trigger the problem but it
>    shows more clearly what is going on later)
> 5. Select the magnifier tool and change to 202% zoom.
> Here, this causes a black background the size of the ellipse's
> bounding box to appear underneath the ellipse. If additional objects
> are drawn the background extends to encompass their bounding boxes. If
> the zoom percentage is changed to 200%, the problem goes away, but it
> occurs at many other zoom levels, too, though usually "even" levels
> like 500% work fine.
> For instance:
> 310%     works fine
> 311%       goes wrong
> 312%-314%  go wrong
> 315%     works fine
> 316%-318%  go wrong
> 319%     works fine
> 320%       goes wrong
> The Double buffering and Background rendering entries in the Debug
> menu do not have any effect, by the way.
> I have an old revision around (609), which works fine in this respect,
> so the error has been introduced somewhere inbetween.
> Martin