From news-rocq.inria.fr!univ-lyon1.fr!pasteur.fr!isdnet!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!news.new-york.net!news.stormking.com!haskin@ptway.com Thu Sep 10 16:30:57 1998 Article: 9605 of rec.games.corewar Path: news-rocq.inria.fr!univ-lyon1.fr!pasteur.fr!isdnet!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!news.new-york.net!news.stormking.com!haskin@ptway.com From: "Brian Haskin" Newsgroups: rec.games.corewar Subject: Re: multiwarrior Date: 9 Sep 1998 19:31:37 -0400 Organization: Storm King Ind. Inc. Lines: 73 Sender: server@news.stormking.com Distribution: world Message-ID: <001f01bddb35$667a90a0$0b00000a@brianjr.haskin.org> Reply-To: haskin@ptway.com NNTP-Posting-Host: valhalla.stormking.com Mime-Version: 1.0 Originator: corewar-l@stormking.com X-Mailer: Microsoft Outlook Express 4.72.3110.5 Hmm, I'm using the win95 version that was released a while back and I'm getting sole wins so it's not a bug in the version I have at least. I'm getting those wins with 9 of Compudemon's Compubunny running and a simple booted core clear (booted because pmars layout of warriors and Compubunny's subsequent distribution seem fairly clumpy). Actually if you want to test it here's my warrior and you can get Compubunny from the warrior archive. ;redcode-94m ;name Shot in the dark ;author Brian Haskin ;strategy clear with all our might ;assert 1 org bt BDIST equ -300 bt mov lst, BDIST for 5 mov {bt, ptr mov.i *ptr, >ptr lst jmp clr, >ptr-2 end Brian Haskin haskin@ptway.com -----Original Message----- From: Robert Macrae To: Multiple recipients of list Date: Tuesday, September 08, 1998 7:02 AM Subject: Re: multiwarrior >John K. Lewis wrote: > >> ./pmars.server harmony diehard impfinity impfinity impfinity >> impish impfinity impfinity impfinity impfinity -r 1000 -b > >> Harmony by P.Kline scores 1705 >> Results: 0 8 0 3 5 6 7 22 38 30 881 >.. >> Not conclusive, but very telling. > >Remarkable performance for Harmony; a further-tuned warrior must be >capable of making it to sole #1 if a stock version can manage 0.8% dual >wins. > >In fact I rather suspect this result illustrates a bug in PMARS. I have >not managed to produce *any* sole winners in a multiwarrior contests, >even with only aggressive warriors. If you run in debug mode, the rounds >seem to end as soon as the third from last warrior dies, leaving two >apparent winners not one? I am using the 286 executable under W95, but >it may be general. > >Anyone else care to check? > >-- > >Regards, > >Robert Macrae > >