FlyHighLittleButterfly

Author

Mue and PEZ

Extends

TeamRobot (although nothing about it is a team robot I think.)

What's special about it?

It wields the gun of Ascendant and runs on the wheels of CassiusClay (yeah, the #1 and #2 bots in the RoboRumble@Home).

Great, I want to try it. Where can I download it?

http://www.robocoderepository.com/BotDetail.jsp?id=2512

How competitive is it?

Expect it to challenge Ascendant itself for the #1 spot. Though regardless if it does take that spot or not this bot will be removed from the rumble when it has proven what it's worth.

How does it move?

CassiusClay/Butterfly (WaveSurfing)

How does it fire?

Ascendant's gun (GuessFactorTargeting with a special treat for WaveSurfers added)

How does the melee strategy differ from one-on-one strategy?

Melee? What's that?

What does it save between rounds and matches?

Nothing is saved between matches. Between rounds all sorts of data is saved both for the movement and for the targeting.

Where did you get the name?

Urging an ascending butterfly to ascend really high.

Can I use your code?

The CassiusClay part, yes. Just read the RWPCL before reading the code. For some reasons beyond my understanding you can't use the Ascendant part of the code though. (PEZ's remark).

What's next for your robot?

Nothing much probably. It's just a test.

Does it have any WhiteWhales?

Shadow naturally. Nothing seems to be able to touch that bot!

What other robot(s) is it based on?

Have you been reading this page at all? =) Ascendant and CassiusClay.

Comments, questions, feedback:

I seriously considered releasing the source code. But the result of my recent test (see comments to Ascendant) changed my mind, because i'd just be providing the source code for a simple gun that beats the crap out of Ascendant. When this major weakness is fixed i'll think again about open sourcing the code (though fixing might take a while due to a lack of time and no real idea on how to fix that right now). --Mue

Cool! A note about this plugging of your gun. I'm not calling doGun() every tick. I only call it from onScanned..() because that's where I call execute(). My radar is stable though so it shouldn't make too much difference. -- PEZ

Yeah, it probably doesnt matter then. I know that A's simple radar only misses very few ticks (number of ticks without scanning the target), and its probably the same with CC's radar. I have to admit though that calling execute() in an event handler is somewhat strange ... to me at least (i'd have expected some kind of stack overflow then, because onScanned is called almost every tick). --Mue

It's just a silly leftover from when CC was a MiniBot. It doesn't have a data manager. I probably should fix that and give my gun and movement that little tiny extra of accuracy. -- PEZ

I probably misunderstood you. I took a look at the code and you dont call execute() in the event handler. You do the gun- and movement stuff there, but execute() is called in run(). Maybe its strange to use the event handlers like that, but i understand now that it works. --Mue

Yeah, it was actually I who were confused. Still am actually. What I mean is I don't call doGun() each tick. The reasons are not those I first babled about. It's because you wanted the movement to run before the gun. -- PEZ

Ah ok. I always thought that the event handlers are called before execute() returns. But anyway, calling doGun in the event handler shouldnt have a measurable effect on bot performance. With that being said it seems that our targetings are on par with each other. So the difference must come from movement. Somewhat unexpected for me, but of course nice to see that my movement is competitive after all (at least if not used vs Shadows or Ascendants gun :-) --Mue

FlyHighLittleButterfly