From news-rocq!jussieu.fr!pasteur.fr!oleane!freenix!news-feed.riddles.org.uk!newsfeed.direct.ca!newspeer.monmouth.com!nntp2.deja.com!nnrp1.deja.com!not-for-mail Fri Jul 30 20:11:13 1999
Article: 10477 of rec.games.corewar
Path: news-rocq!jussieu.fr!pasteur.fr!oleane!freenix!news-feed.riddles.org.uk!newsfeed.direct.ca!newspeer.monmouth.com!nntp2.deja.com!nnrp1.deja.com!not-for-mail
From: Ben Ford <fordbc@hotmail.com>
Newsgroups: rec.games.corewar
Subject: 0.66c WingShot
Date: Fri, 30 Jul 1999 14:50:10 GMT
Organization: Deja.com - Share what you know. Learn what you don't.
Lines: 56
Message-ID: <7nse2u$h40$1@nnrp1.deja.com>
NNTP-Posting-Host: 207.170.76.162
X-Article-Creation-Date: Fri Jul 30 14:50:10 1999 GMT
X-Http-User-Agent: Mozilla/4.5 [en] (WinNT; U)
X-Http-Proxy: 1.0 x34.deja.com:80 (Squid/1.1.22) for client 207.170.76.162
X-MyDeja-Info: XMYDJUIDbenford
Xref: news-rocq rec.games.corewar:10477

Here is my 0.66c version of WingShot.  It doesn't score much differently
than the 0.8c version.  Its one instruction smaller and has slightly
better seperation between scan locations.  In theory the step size of
this version should be able to be modified to a better multiple of 10
but my tests show it scoring worse whenever I try it.
I belive the biggest improvements WingShot made over BiShot was the the
slight off-axis scanning and using a djn to automatically go into the
coreclear if nothing is found after two passes through the core.  A
couple of other improvements are to be found in the code if you care to
look for them.
This version of WingShot is on the big (size 55440) hill at koth.org.
For some reason it scored several points higher than the 0.8c version
there.

-Ben

;redcode-94
;name WingShot 0.66
;author Ben Ford
;strategy 0.66c BiShot-style oneshot
;assert !(CORESIZE%10) && (MAXLENGTH>=20)

start	equ	oscn

oinc	equ	10
oset	equ	(10*((MAXLENGTH+MINDISTANCE)/10)+1)
odif	equ	(10*(CORESIZE/20))
otic	equ	(CORESIZE/10)

ofnd	add.f	 oadd,	 optr
oscn	sne.i	*optr,	@optr
	djn	 ofnd,	#otic
	djn.f	 ospl,	@oscn
oadd	dat	#oinc,	#oinc

for	10*((MAXLENGTH-20)/10)
	dat	0,	0
rof

	dat	0,	0	; scanned hi
	dat	0,	0
	dat	0,	0
optr	spl	{oset,	>odif+oset-4
	dat	0,	0	; scanned lo

	dat	{10,	>odif
ospl	spl	#10,	<-MINDISTANCE
oclr	mov	ospl,	}optr
	mov	ospl,	>optr
obmb	djn.a	oclr,	<ospl

end	start


Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.