On Apr 30, 2005, at 10:26 AM, Dmitry Markman wrote:
we have very complex molecular dynamic simulations
and I can not see any thing close to 10% of 1.3.1 slowness
Java 1.4.2 is very _very_ slow at drawing.
Well look at the following results when using the attached code which
draws 1 million lines (times in milliseconds)...
(PM G5 Dual 2GHz, 6800 GT with 256MB, and dual 23" displays running
at 1920x1200x32)
With Quartz 2D Extreme disabled (the default for 10.4)...
[Session started at 2005-04-30 11:12:40 -0700.]
Java Version:1.4.2_07
OS Version:10.4
8899
8499
8688
8914
With Quartz 2D Extreme enabled...
[Session started at 2005-04-30 11:13:46 -0700.]
Java Version:1.4.2_07
OS Version:10.4
466
308
308
316
311
Note I modified your original code to move the random number
calculation used to pick points to outside of the drawing loop which
is being measured. Also I commented out all of you hint, etc. stuff.
final JFrame f = new JFrame("Example");
f.setSize(400,400);
JButton b = new JButton("Press Me");
f.getContentPane().add(b,BorderLayout.NORTH);
f.show();
b.addActionListener(new ActionListener()
{
public void actionPerformed(ActionEvent e)
{
Random r = new Random();
Graphics2D g = (Graphics2D)(f.getContentPane
().getGraphics());
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Java-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/java-dev/email@hidden
This email sent to email@hidden