Scratch

Okay, after much experimenting I found a way to get it to work...

What I did was in the Stage scripts to change "broadcast MoveBall"
to this block:
"broadcast MoveBall and wait"

I'm not sure why yet, but that worked for me. Could you please try that and let me know if it also works for you?

The other thing I did was to remove all the "stop script" blocks, since they aren't needed at the bottom of a stack (only if you want to interrupt a script in progress). However, that probably shouldn't make a difference. (Note: I left on the "stop all" blocks, which are important.)

I really like the interaction and sound effects for your game!

I tried using "Broadcast MoveBall and Wait", but now it runs painfully slow...

http://scratch.mit.edu/projects/PatternsOfChaos/7986

Last edited by PatternsOfChaos (2007-05-29 01:28:34)

In the meantime, try this: after "broadcast MoveBall" add a "wait 0.0001 seconds" that also seems to work.

But I'll ask others here why the "forever if "broadcast"" without a wait doesn't work in the Java player.

Thanks all, I thought I was going crazy, because I had the same problem....
http://scratch.mit.edu/projects/twoPi/6402
This thing worked great on the standalone machine, but froze on the browser.  Subsequent tests (motionTest 1 - 4) prove that the problem is with the broadcast message call. 

If the broadcast is happening within a loop and the loop has a wait in it (the normal circumstance) that should do the job. 

Frog lunch 1.1

http://scratch.mit.edu/projects/twoPi/8158

shows this technique.

It still is a bit too slow. I don't know how much is the java engine and how much is the built-in delay. It was a touch slower on standalone, too, but .001 seconds shouldn't be any worse than the delay you'll already get from java. I can't imagine it's running at 100 frames per second. 

I am seeing an occasional lag.  Could be system resources are building up. Does the Java applet create a new Thread for each forever loop?

'Invalid Request' usually means that the project has been removed.

http://scratch.mit.edu/projects/PatternsOfChaos/8270

Well, I got it working with a delay in the loop, but this time the ball seems to pass through the paddle. Ball_TestCollide is never broadcast

Best I can tell, the loop executes too quickly for everything to be called on every loop, and the delay either makes it too slow if the increment is small, or the increment is too large - in either case, the collision test fails.

I'd appreciate any other ideas on this - but it seems that the best course is looking at optimization of the MoveBall funcitons...

Hi, everyone.

I just fixed a subtle bug in the Java player that I think was responsible for many of the problems reported here. Those of you who worked around the problem by adding extra wait commands might try removing those wait commands.

Please let me know if you see any more cases where broadcasts are ignored in the Java player but not in Scratch.

Thanks for your patience.

  -- John

P.S. You may need to quit and re-start your browser to get the new player. You are running the new player if you see "v13" in tiny letters in the upper-left corner of the Player. The old one was "v12".