Path: utzoo!attcan!uunet!decwrl!sdd.hp.com!ucsd!ames!dftsrv!amarna.gsfc.nasa.gov!xrtnt
From: xrtnt@amarna.gsfc.nasa.gov (Nigel Tzeng)
Newsgroups: comp.sys.amiga.games
Subject: Re: crobots
Summary: maybe comp.sys.amiga.games.crobots?  Nah, too hard to type
Message-ID: <2895@dftsrv.gsfc.nasa.gov>
Date: 25 Jul 90 14:10:46 GMT
References: <2842@dftsrv.gsfc.nasa.gov> <1626@nyx.UUCP> <9007220151.AA02222@melmac.umd.edu> <1654@nyx.UUCP>
Sender: news@dftsrv.gsfc.nasa.gov
Reply-To: xrtnt@amarna.gsfc.nasa.gov
Distribution: na
Organization: NASA Goddard Space Flight Center - Greenbelt, MD, USA
Lines: 43
News-Software: VAX/VMS VNEWS 1.3-4


  In article <1654@nyx.UUCP>, you write...
  ^ 
  ^As for Berserker (now at 4.0 and, like a Microsoft program, getting slower
  ^all the time) I'm on the verge of junking it in favor of a total rewrite.  It
  ^is still an absolutely crack shot - it can hit just about anything, no matter
  ^what it's doing, unless it actually changes course after the shot is released,
  ^but it's only capable of firing about every 10 seconds at best... when the
  ^target is moving fast, maybe every minute or so... it still HITS, really makes
  ^some beauty shots from long distances, but that does no good if it can be killed
  ^between one shot and the next.  The whole program needs some serious help, and
  ^optimization (at least the way *I* do it...) just makes it worse.
  ^ 
  ^I wish someone would post SOURCE CODE for their robots - the last two which
  ^were posted here were very very good programs, even though they're rotten 
  ^shots and aren't real fancy movers either... but they win.  Annoying.
  ^ 
  ^..                           <<<<Infinite K>>>>
  ^ 
  ^..---------------------------------------------------------------------------.
  ^|Ben Scott, professional goof-off and consultant at The Raster Image, Denver|
  ^`---------------------------------------------------------------------------'

  I'll upload mine and either post it or mail it to you.  I'd like to see your
  tracking logic before you junk it ;-).  I had a fairly elaborate one in a later
  robot (now lost) but it had problems with identification in the "melee" field. 
  Crossover paths would confuse it a bit...

  You should probably opt for a 'racetrack' type of movement.  Lousy trackers
  have bad times with leading unless they program for one type of robot (IMHO 
  that's cheating ;-).  If you can cut it close enough to the wall you reduce 
  your scanning area by quite a lot...  That WAS the original intent of such a
  path but the lousy tracking logic I first put in really didn't take advantage
  of this.  Plus it is still a little too far out from the wall for me to be
  really comfortable about not scaning supposed "empty" areas ;-).

  Nigel

  ------------------------------------------------------------------------------
     // | Nigel Tzeng - STX Inc - NASA/GSFC COBE Project
   \X/  | xrtnt@amarna.gsfc.nasa.gov
        | 
  Amiga | Standard Disclaimer Applies:  The opinions expressed are my own.