From news-rocq.inria.fr!jussieu.fr!math.ohio-state.edu!howland.erols.net!news-peer.gsl.net!news.gsl.net!hammer.uoregon.edu!arclight.uoregon.edu!enews.sgi.com!insync!news.azstarnet.com!news Tue Oct 22 11:24:52 1996 Article: 6336 of rec.games.corewar Path: news-rocq.inria.fr!jussieu.fr!math.ohio-state.edu!howland.erols.net!news-peer.gsl.net!news.gsl.net!hammer.uoregon.edu!arclight.uoregon.edu!enews.sgi.com!insync!news.azstarnet.com!news From: Franz Newsgroups: rec.games.corewar Subject: MoreVamp v4 Date: Tue, 22 Oct 1996 01:32:03 -0700 Organization: Starnet Lines: 64 Message-ID: <326C8683.4339226D@azstarnet.com> NNTP-Posting-Host: usr6ip44.azstarnet.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 3.0 (X11; I; Linux 1.2.13 i586) This is as far as I got with this warrior along the lines of the original ... any other change would probably change the general idea behind it ... i might do that .. been meaning to doa good vampire ... the one thing that's ugly now is that it uses stepsize 5 .. oh well ... the detect location which detects own fangs/papers/imps is now covered with spl's to kill if a paper triggered the coreclear ... the switch to coreclear is still painfully slow so most things are out of tehre before it gets a chance to fry them ... it lowered the overall wilkies benchmark by 2 points but added against papers a bit ... I will work on that now and on the bombing itself (it should be optima or bigger number and a shift or something) ok here is the MoreVamp v4 .. still this gives 89 wilkies so not a strong warrior ... definatelly a LOT of room to improve and change .... ;redcode-94 ;name MoreVamp v4 ;author Franz ;start vampire ;assert (CORESIZE%5)==0 step equ 5 detect equ 200 org vamp ;lets start at 2000 since that should be covered with the pit ccp dat pit,fang-detect-20 dat 0,0 gate dat 0,0 dat 0,0 fang jmp pit-100, #100 dat 0,0 ;i'll ad self splitting to enhance the durability ; vamp spl #-step, #step incr add.f vamp, fang mov.i fang, @fang ;this will give us not only detection of our fangs ;coming in but also of imps! and the clear is fairly anti imp ;not the best but worth the effort:) jmz.f incr, fang-detect mov.i gate,>kill ;standard spl/dat coreclear cc spl #sec-ccp, #60 mov *ccp, >ccp jmp -1,>gate ;lets put the pit on the bottom so we can have it carpet bomb ;and to optimize a bit we'll put the stepsize on the first spl ;to avoid the dat for 50 dat 0,0 rof pit spl #cc-ccp,>2000 sec spl #2, #30 mov -1,<2000 kill jmp -3, >gate end