Difference between revisions of "User talk:Staticline"
Staticline (talk | contribs) |
Staticline (talk | contribs) |
||
| (3 intermediate revisions by 3 users not shown) | |||
| Line 2: | Line 2: | ||
I never had time to take a deeper look into Jython, because of time and possible compability reasons. Currently I'm stuck to Java because my tutor wants me to have the Code in Java. I'll take a look into Jython and Robocode after my machine learning course when I work on my own :) --[[User:Staticline|Staticline]] 16:03, 24 June 2010 (UTC) | I never had time to take a deeper look into Jython, because of time and possible compability reasons. Currently I'm stuck to Java because my tutor wants me to have the Code in Java. I'll take a look into Jython and Robocode after my machine learning course when I work on my own :) --[[User:Staticline|Staticline]] 16:03, 24 June 2010 (UTC) | ||
| + | |||
| + | About the Jython overhead, while I'm not a fan of microbenchmarks, [http://blog.dhananjaynene.com/2008/07/performance-comparison-c-java-python-ruby-jython-jruby-groovy/ this gives a rough idea]. In other words, Python in this test is at least 120x slower than Java, and Jython is over 3x slower than Python... so... I'd guess that while one might be able to make a Jython bot perhaps, it's capabilities would either be a bit limited, or require offloading the tight loops to java code. --[[User:Rednaxela|Rednaxela]] 23:25, 24 June 2010 (UTC) | ||
| + | |||
| + | Maybe I should try to use some fancy C-libraries with a Python overhead running running in Jython :) Well, first I have to improve my bot algorithms. Fast or slow running crap is still crap. --[[User:Staticline|Staticline]] 11:00, 26 June 2010 (UTC) | ||
| + | |||
| + | I'm afraid it's my fault that your bot is ranking that low. On my client you score zero, because your bots gets disabled, see below. Sorry, I'll stop my client until the cause is known. --[[User:GrubbmGait|GrubbmGait]] 18:11, 26 June 2010 (UTC) | ||
| + | <pre>SYSTEM: An error occurred during initialization of staticline.Whiskey 0.5.1 | ||
| + | SYSTEM: java.lang.NoClassDefFoundError: Could not initialize class staticline.Whiskey | ||
| + | java.lang.NoClassDefFoundError: Could not initialize class staticline.Whiskey | ||
| + | at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method) | ||
| + | at sun.reflect.NativeConstructorAccessorImpl.newInstance(Unknown Source) | ||
| + | at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(Unknown Source) | ||
| + | at java.lang.reflect.Constructor.newInstance(Unknown Source) | ||
| + | at java.lang.Class.newInstance0(Unknown Source) | ||
| + | at java.lang.Class.newInstance(Unknown Source) | ||
| + | at robocode.battle.Battle.unsafeLoadRobots(Unknown Source) | ||
| + | at robocode.battle.Battle$UnsafeLoadRobotsThread.run(Unknown Source) | ||
| + | </pre> | ||
| + | |||
| + | Hey. Sorry for that trouble. I had some minor problems with my 0.5 bot, fixed them in 0.5.1 and thought the problems were solved. Obviously I was wrong - thanks for your info. I'm removing this version from the participants list and do some more tests/fixes next week when I'm back at home (and back on my production machine). --[[User:Staticline|Staticline]] 23:16, 26 June 2010 (UTC) | ||
Latest revision as of 00:16, 27 June 2010
Welcome to Robowiki. It is great to see so many new users joining our community. As for a python version, if there is a version of python that works with Java (see Jython), it might be possible. But I am unsure what the overhead would be! — Chase-san 14:51, 24 June 2010 (UTC)
I never had time to take a deeper look into Jython, because of time and possible compability reasons. Currently I'm stuck to Java because my tutor wants me to have the Code in Java. I'll take a look into Jython and Robocode after my machine learning course when I work on my own :) --Staticline 16:03, 24 June 2010 (UTC)
About the Jython overhead, while I'm not a fan of microbenchmarks, this gives a rough idea. In other words, Python in this test is at least 120x slower than Java, and Jython is over 3x slower than Python... so... I'd guess that while one might be able to make a Jython bot perhaps, it's capabilities would either be a bit limited, or require offloading the tight loops to java code. --Rednaxela 23:25, 24 June 2010 (UTC)
Maybe I should try to use some fancy C-libraries with a Python overhead running running in Jython :) Well, first I have to improve my bot algorithms. Fast or slow running crap is still crap. --Staticline 11:00, 26 June 2010 (UTC)
I'm afraid it's my fault that your bot is ranking that low. On my client you score zero, because your bots gets disabled, see below. Sorry, I'll stop my client until the cause is known. --GrubbmGait 18:11, 26 June 2010 (UTC)
SYSTEM: An error occurred during initialization of staticline.Whiskey 0.5.1 SYSTEM: java.lang.NoClassDefFoundError: Could not initialize class staticline.Whiskey java.lang.NoClassDefFoundError: Could not initialize class staticline.Whiskey at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method) at sun.reflect.NativeConstructorAccessorImpl.newInstance(Unknown Source) at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(Unknown Source) at java.lang.reflect.Constructor.newInstance(Unknown Source) at java.lang.Class.newInstance0(Unknown Source) at java.lang.Class.newInstance(Unknown Source) at robocode.battle.Battle.unsafeLoadRobots(Unknown Source) at robocode.battle.Battle$UnsafeLoadRobotsThread.run(Unknown Source)
Hey. Sorry for that trouble. I had some minor problems with my 0.5 bot, fixed them in 0.5.1 and thought the problems were solved. Obviously I was wrong - thanks for your info. I'm removing this version from the participants list and do some more tests/fixes next week when I'm back at home (and back on my production machine). --Staticline 23:16, 26 June 2010 (UTC)
