http://members.optusnet.com.au/ksaho/work/test.jar
I don’t have access to any computer slower than my own currently, so I want you guys to tell me how fast/slow my app is.
Press the “A” button to start the animation.
On my PC this app runs without any performance issues.
Then again I have a A64 3K+ with 1GB RAM.
You provided the source files… This means we need to compile them laboriously?
You should compile your test and put into an executable JAR for example, so that we testers can test it by the matter of one mouse click…
Since Java is platform neutral it should run on any Java enabled platform then.
{Edit}
You don’t want us to test the performance of your lame “test”. You really exploit the good-naturedness of some people…
I went into the trouble of importing the sources in eclipse, but it won’t compile (got 33 errors).
Maybe bombadil put it a bit harsh, but he’s basically right.
You can’t expect people to find out themselves what to do with the resources you provide without any explanation and receive valuable feedback. Just make it as easy as possible to download and run, ie at least something like “download the jar at http://bla.com/app.jar and type java -cp app.jar main.MainClass”, or even better, webstart.
Thanks.
I will have to learn to use Webstart, for now I will just go back to using Eclipse.
I figured it would be easy to just throw the sources into an IDE and compile.
Sorry about this, I will use Eclipse to put everything into a single file and re-upload this.
I need to know if I should scrap my current plan of using Java2D + Swing or should I just go to Java3D/Xith3D with Swing.
I don’t have as much experience as most of you guys do and from time to time I will ask for help. I tend to assume that you guys know how to do anything and everything.
FYI, you can create a jar from the command line as well and you don’t have to switch IDE’s just for creating a jar.
Go to the base directory of where your .class files are (this should be the directory directly above your your ‘display’, ‘entities’, ‘game’ etc directories which hold your compiled classes).
Create a file called main.txt with the following line in it (assuming your main class is main.Main as I believe it is):
Main-Class: main.Main
(don’t forget to have a CRLF at the end of this line)
type:
jar cmf main.txt someapp.jar *
This should create an executable jar.
[quote]type java -cp app.jar main.MainClass", or even better, webstart.
[/quote]
If name of application isn’t main and name of package isn.t com.main then I’m strictly against it. It has been a nice tradition to start an application by class with the same name as the application, or by a class named ApplicationNameLauncher.
It look like sentence: “He seen seen seen some action.”
Doubleclickable jar, or cmd/sh file is enought.
Well… it works… but what is it supposed to do ?
All i have is a ‘collision’ message as soon as i click on ‘A’ and no animation.
And something about Mickael J when i click on the… circles 
Is it supposed to be animated ?
tested on XP/3.2Ghz and Linux 800Mhz (jdk1.5)
Lilian
[quote]I create an unfinished swing based system which I’m entirely unsure of which I should or shouldn’t continue based on general performance.
[/quote]
Oh well, I did compile your classes in order to help, right before I edited my original message!
What did I see as “tester of performance” ?
What’s the thread’s name? How’s the performance so far?
Well, so what did the “tester” see when he dared to start your “unfinished product” ? I’d to click with the mouse somewhere and got an infantile and offending message printed out as result. That’s all! There’s no single bit in your “test” to judge any performance! So how should we tell you that your test worked well on our PCs with lower config than your monster PC?
I take it you fooled us. I’m not here in this forum to get fooled. I want to learn and help with technical questions about Java gaming. This is no kindergarden.
… then I edited my original post and said you did not want us to test performance.
It shouldn’t pause everytime an event occurs.
Your mouse shouldn’t lag when moving it across the screen.
Memory consumption should be under 30MB.
These are problems lower end machines have had previously with my programs.
I have erased the silly message.
Anyway, you’re saying that if I were to ask for help about performance of my program that it should be done elsewhere because this forum is reserved only for technical discussions?
[quote]It shouldn’t pause everytime an event occurs.
Your mouse shouldn’t lag when moving it across the screen.
Memory consumption should be under 30MB.
These are problems lower end machines have had previously with my programs.
[/quote]
This is what you should have been saying in the first article, so that everybody willing to test knows what he should look at. Together with a compiled sample/JAR (Netbeans does it for you when you provide a Manifest I think.)
However you also need to include performance parameter printings in your program (printf’s like “Hello you!” don’t indicated anything).
For example the needed memory you can query and print out.
[quote]Anyway, you’re saying that if I were to ask for help about performance of my program that it should be done elsewhere because this forum is reserved only for technical discussions?
[/quote]
No. Performance disussions do belong to this forum, naturally. However your original article and “test” had nothing to do with performance or at least it was totally unclear what the tester should have done…
Your latest article goes into the right direction.
A good start to see what performance measuring is about and how to make tests so other people could help you, is Jeff’s book:
http://java.sun.com/docs/books/performance/
Thank you, sorry about the merrygoround.
I guess my manners need a little working.
Performance is an area of which I’m paranoid about.
My programs usability reflects upon it’s performance and poor performance usually sends the wrong message about code and/or my understanding of the language.
Performance wise, what kind of an effect does Swing have on it’s own in any program?
Swing is fast, but AWT’s event queue can get filled up pretty fast which could derive poor performance.
I wish to use Swing for high performance applications because of Swing’s ease of use and low maintenance costs.
The AWT queue only gets filled up if you are doing something wrong. If you need to do an extended operation upon recieving an event, handle it in another thread.
[quote]Performance is an area of which I’m paranoid about.
[/quote]
Yes, I’ve noticed that