From news-rocq.inria.fr!univ-lyon1.fr!newsfeed.nacamar.de!wnfeed!worldnet.att.net!4.1.16.34!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!news.new-york.net!news.stormking.com!iltzu@sci.fi Tue Sep 30 09:51:16 1997
Article: 7872 of rec.games.corewar
Path: news-rocq.inria.fr!univ-lyon1.fr!newsfeed.nacamar.de!wnfeed!worldnet.att.net!4.1.16.34!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!news.new-york.net!news.stormking.com!iltzu@sci.fi
From: iltzu@sci.fi (Ilmari Karonen)
Newsgroups: rec.games.corewar
Subject: Shelter
Date: 29 Sep 1997 16:49:59 -0400
Organization: Storm King Ind. Inc.
Lines: 66
Sender: server@news.stormking.com
Distribution: world
Message-ID: <29199740.Amiga@pop.sci.fi>
Reply-To: iltzu@sci.fi
NNTP-Posting-Host: valhalla.stormking.com
Originator: corewar-l@stormking.com
X-Mailer: Air Mail V2.0 (Unregistered) -- Amiga Mailer by Danny Y. Wong

It's been about a month or something now since Anton Marsden said he'd
post all the tournament entries soon, and since there's been discussion
about "silver bullets", I figured I'd post my round 2 entry, Shelter:

It exploits two weaknesses of Electric Head's mini HSA:
1. A decoy, placed just before the program, can render it invisible to
   HSA.
2. The HSA degenerates into a .66c forward core clear when it hits a
   decoy. This means that if it starts wiping the trail of a faster
   clear, such as d-clear, it's doomed.

This version is a result of much trial-and-error tweaking, as one can
see from the clear. Believe it or not, this is the most effective size.

;redcode-94
;name Shelter 1f3
;author Ilmari Karonen
;strategy Anton Marsden's tournament round 2 entry:
;strategy  hide from HSA behind a decoy, and catch
;strategy  it with a brainwashing 1c forward clear
;date Jul 31, 1997
;planar clear, pspace
;assert CORESIZE == 8000

PSTATE  equ     250     ; from EH
STATES  equ     10      ; from EH
BWASH   equ     2

hsastep equ     9       ; from EH's mini HSA
proglen equ     7

cend    equ     (cptr+BWASH+2*STATES)

        org     start

decoy   z for   (MAXLENGTH-19)
        spl     #-z, 4000+z
        rof

dgap    dat.f   0, 0    ; next scan will land at dgap+hsastep+1
        dat.f   0, 0    ; or at dgap+hsastep+2
        dat.f   0, 0
cptr    dat.f   0, 0
        dat.f   0, 0
        dat.f   0, 0
        dat.f   0, 0
        dat.f   0, 0
cbmb    dat.f   {1234, {BWASH+2*STATES
cwsh    stp.ba  #250, #BWASH+2*STATES
        dat.f   0, 0    ; this will be scanned
        dat.f   0, 0    ; or this

start   spl     #0, 0
loop    mov.i   cbmb, >cptr
        mov.i   cwsh, >cptr
        mov.i   cwsh, >cptr
        mov.i   cwsh, >cptr
        djn.f   loop, >cptr

        end

--
Ilmari Karonen (iltzu@sci.fi)
http://www.sci.fi/~iltzu/