--Joe Bowbeer wrote:
Is any progress being made with bug 4151279?
http://developer.java.sun.com/developer/bugParade/bugs/4151279.html
Is this scheduled to be fixed in Tiger 1.5?
Is this the right place to ask about this?
This is the best place I can think of to ask this question.
This is a complicated issue whose full and complete solution will definitely exceed the Tiger timeframe as it would involve finding agreement between a lot of different internal rendering algorithms.
But I think we can address at least one known "really bad offender" in the Tiger time frame which is the fact that our software renderer maintains subpixel precision when filling arcs, ovals, and Shapes but rounds to integers when drawing their (thin) outlines. This not only creates inconsistencies between the fills and the draws, but it means that our draws are fairly angular and crooked.
Some of the things that make it complicated are:
- Some of the bug reports talk about fillOval and others
talk about drawOval, yet customers that add their voices
to such reports don't always distinguish their problems
- The comment about "good enough" in the bug you cited
was specifically referring to filled ovals and those,
by and large, do appear fairly consistent and symmetric
although many of the comments will include test cases
based on drawOval and claim we are wrong.
- Internally there are many pathways that we might take
to satisfy the same rendering request depending on
which attributes you set:
- We might use X11 or GDI
- We might use software loops
- We might express arcs and ovals as Cubic beziers
- We might draw outlines using bresenham
- We might draw outlines by converting to
a "Stroke Shape" and filling it
- We might use different pipelines for
filling vs. drawing depending on the
attributes specifiedI could share some anecdotes of our experiences with some of the platform API's attempts to draw arcs, but basically we aren't always the bad guys here and we have to discover their problems before we can work around them.
Hopefully that answer helped. If you have a specific test case that you are specifically worried about, it might help to identify it (using our comments web form on java.sun.com) rather than cite that bug report (which talks about both filled and drawn ovals)...
...jim
=========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA2D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
