From news-rocq.inria.fr!jussieu.fr!fu-berlin.de!news-peer.gip.net!news.gsl.net!gip.net!newshub.northeast.verio.net!nntp.upenn.edu!msunews!mooredav Mon Apr 20 21:29:04 1998 Article: 8890 of rec.games.corewar Path: news-rocq.inria.fr!jussieu.fr!fu-berlin.de!news-peer.gip.net!news.gsl.net!gip.net!newshub.northeast.verio.net!nntp.upenn.edu!msunews!mooredav From: mooredav@cps.msu.edu (David Matthew Moore) Newsgroups: rec.games.corewar Subject: Re: Olympic report 1998-04-18 Date: 20 Apr 1998 19:01:41 GMT Organization: Michigan State University Lines: 53 Message-ID: <6hg62l$s7l$1@msunews.cl.msu.edu> References: <19980414015707.13891.rocketmail@send1b.yahoomail.com> <6h8d0a$lrl@news-rocq.inria.fr> <35380881.354A@dial.pipex.com> <3536B6C5.3ADC@dial.pipex.com> <6hafqj$55i@news-rocq.inria.fr> <35398ADE.5D3B@dial.pipex.com> NNTP-Posting-Host: umnak.cps.msu.edu X-Newsreader: TIN [version 1.2 PL2] Robert Macrae (ff95@dial.pipex.com) wrote: : Planar wrote: : > >>> 650 *RedRain Derivative 31/ 51/ 19 110 110.41 : > : > Note how this minimal scanner absolutely destroys CLP (97.3% wins). : > It seems to be built for this purpose. RedRain is of course helpless : > against imps (it has no gate) and papers (it doesn't stun). Also, the : > score against elf is interesting (100.0% ties). : I did have it in mind as a P-component, but I have not found a role for : it as yet. As I remember the scan-pattern is mod-4 so my version will be : terribly inneffective against bombers or clears which include an SPL. Check out this little bomber, also written by Wayne Sheppard: ;Name No ties allowed mov <21,1+2234 sub 1,-1 jmp -2,-2234 Doesn't this look a lot like the stones from CW65? It's a mod-2 bombing pattern, followed by a subtraction core clear. Back in the days of 88, we didn't have: SPL #0, 0 we used: SPL 0, 0 so you could subtract from this and get: SPL -123, 456 so subtraction core-clears were nearly as lethal as MOV ones! This was used in many self-mutating programs including Twill and my own PacMan and Freight Train. After trashing every core location, this little bomber mutates into a JMP 0 instruction and waits for the end of time. Nice! But it gets even better! If Wayne had used predecrement indirect < on that last instruction, then you get a little pop-up gate at the end! It's a harmless gate (JMP 0, <-1) but at least it doesn't lose so much versus imps. ;Name No ties allowed 2 mov <21,1+2234 sub 1,-1 jmp -2, < -2234