Client java version

Fragment of a discussion from Talk:RoboRumble
Jump to navigation Jump to search

I do not think that above report indicates that MoxieBot is broken. Actually, the log indicates that the bot was working and was reporting sensible things. The part about

SYSTEM: ncj.MoxieBot 1.0 has died

is due to its logic: MoxieBot does not fire when there are no bullets in the air. If the other bot runs out of energy, then they both wait and robocode gives an energy penalty kick after certain "no fire" timeout. Eventually they both killed by penalty energy drain, which is reported as "has died".

I played MoxieBot in the GUI it gives similar log entry while it is moving and firing.

Beaming (talk)01:31, 5 September 2017

In the log above, there are 25 lines of "Hit by a wave that was not being correctly tracked", indicating MoxieBot losing energy. That would be consistent with MoxieBot stalling and MyFirstRobot getting a 100% hitrate on it. MyFirstRobot fires 1.0 power bullets, which cause 4.0 damage, and 100 / 4 = 25 hits to kill = 25 log messages.

I also made a video of MoxieBot vs. TrackFire on that computer. It stalls for 2 rounds, but works on the other 8. It doesn't seem so bad when not minimized.

I wonder if the bug is that MoxieBot loses its radar lock and cannot regain it, which wouldn't be that hard of a fix...

MultiplyByZer0 (talk)02:37, 5 September 2017
Edited by author.
Last edit: 04:44, 5 September 2017

Yep. This is what I see as well. It seems to be a bug in the MoxieBot logic. Not in the robocode.

By the way ThreadDeath is also there. Looks like I need to try java7 or earlier and see if ThreadDeath related losses are there.

Beaming (talk)04:37, 5 September 2017

I also checked the historical robocode 1.8.3, MoxieBot has the same buggy behaviour.

Beaming (talk)04:41, 5 September 2017