[Home]SmartFactorTargeting

Robo Home | Changes | Preferences | AllPages

[/OldChat]

Introduction

This is a technique that I am trying to implement in AgentSmith. As I have not finished implementing it I don't know how good or bad it will be. As you can guess from the name its a variation on GuessFactorTargeting but thats as much as I am willing to say! I need all the help I can get against the top bots without giving away some of my secrets!

Once I release the bot into the rumble I will probably release details of the technique! Check back here for updates on how good or bad my implementation is! (Probably going to be really bad!)

--Wolfman

Update 01/03/2006:

Just getting back into writing AgentSmith again after a bit of an absense! I've changed my mind on how to do this technique about three times so after having written it three different ways I think I might (with luck) be starting to get somewhere that is at least reasonable, so long as I can sort out a problem with memory that I seem to be having. Does anyone have any ideas why my bot will be running of memory? Is there any way I can track it to see whats using what etc? (Im using Eclipse by the way but don't know much about it). I've tried removing as many calls to "new" as possible so I just allocate everything at the start of the battle and never do it again. Still it seems to be running out after about 60 rounds ....

Maybe you have some circular reference somewhere so that the garbage collector can't free stuff. I don't think it's a good idea to move those object creations to the start. Better new them as they are needed and let the garbage collecter reap. You probably have a huge structure that you keep adding stuff too. That's where you should start looking. -- PEZ Oh well I guess I will just have to plod along and see how I can improve things! --Wolfman


Robo Home | Changes | Preferences | AllPages
Edit text of this page | View other revisions
Last edited March 1, 2006 13:27 EST by PEZ (diff)
Search: