Hello everybody,
yesterday, I posted about a rendering bug with antialiased bar charts. The best example I could find that exhibits this behaviour was the ParetoChartDemo that comes with the distribution.
Apparently, this is not just an antialiasing bug but a major rendering bug: when I disable antialiasing with the ParetoChartDemo on Mac OS X 10.3 (JDK 1.3.1 and 1.4.2), the bar outlines are drawn in a completely weird way. If nobody has seen this problem before on this or any other platform, we might have a real major bug in the Mac's JDK here. I will also notice Apple about this.
However, I would be grateful if someone over here could take a look at the following page I quickly slapped together for illustration purposes:
http://homepage.mac.com/ralphscheuer/PhotoAlbum7.html
Has anybody seen something like this before?
Thanks for your advice.
Ralph[/url]
JFreeChart rendering bug?
-
- JFreeChart Project Leader
- Posts: 11734
- Joined: Fri Mar 14, 2003 10:29 am
- antibot: No, of course not.
- Contact:
There was some discussion about this on the Java2D mailing list:
http://www.javadesktop.org/forums/threa ... &tstart=15
I haven't looked really closely at it yet...
http://www.javadesktop.org/forums/threa ... &tstart=15
I haven't looked really closely at it yet...
David Gilbert
JFreeChart Project Leader
Read my blog
Support JFree via the Github sponsorship program
JFreeChart Project Leader


-
- Posts: 4
- Joined: Wed Mar 10, 2004 3:37 pm
Hello,
via Apple's java-dev list, I just received the following statement regarding the coordinates used to draw the chart:
Thanks for your help
Ralph[/quote]
via Apple's java-dev list, I just received the following statement regarding the coordinates used to draw the chart:
Is it possible that the coordinates for the charts are fractional? And if they really are, can I force JFreeChart to use Integer coordinates only?If they're fractional, I could imagine getting exactly the behavior you see, as the rendering code tries to correctly draw "virtual" pixels that straddle physical pixel boundaries. The misplacement of "aliased" lines is also plausible, depending on how the fractional coordinates got rounded.
Thanks for your help
Ralph[/quote]