From news-rocq.inria.fr!univ-lyon1.fr!ghost.dsi.unimi.it!batcomputer!hookup!news.kei.com!yeshua.marcam.com!zip.eecs.umich.edu!newsxfer.itd.umich.edu!nntp.cs.ubc.ca!uw-beaver!cs.washington.edu!jayhan Wed Mar 23 16:12:21 1994 Article: 531 of rec.games.corewar Newsgroups: rec.games.corewar Path: news-rocq.inria.fr!univ-lyon1.fr!ghost.dsi.unimi.it!batcomputer!hookup!news.kei.com!yeshua.marcam.com!zip.eecs.umich.edu!newsxfer.itd.umich.edu!nntp.cs.ubc.ca!uw-beaver!cs.washington.edu!jayhan From: jayhan@cs.washington.edu (Jay "Thierry" Han) Subject: Some redcode [Re: koth instructions] (long) In-Reply-To: mconst@soda.berkeley.edu's message of 22 Mar 1994 01:47:30 PST-8 Message-ID: Lines: 577 Sender: news@beaver.cs.washington.edu (USENET News System) Organization: Computer Science & Engineering, U. of Washington References: <19940321134706IZZYVD5@mvs.oac.ucla.edu> <2mlinj$knv@agate.berkeley.edu> Date: 22 Mar 1994 11:57:37 GMT In article <2mlinj$knv@agate.berkeley.edu> mconst@soda.berkeley.edu (Michael Constant) writes: > From: mconst@soda.berkeley.edu (Michael Constant) > Subject: Re: koth instructions... > Date: 22 Mar 1994 01:47:30 PST-8 [...] > successful program (like Imperfection (second place on stormking-94x with > a 27-point lead (!) over third place-- yay!)) because I don't want to have to > resubmit the old version if it turns out to be better. And in any case, Gotcha! Michael, it's about time you published the code for it! Okay, here's my contribution. Thanks to James who unveiled just how silly his Silly was. :-) Here I go. First off, I have to say that the 94x Hill is much milder than any other hill. Also, I love the bigger core! It *does* let you write bigger, more complex programs. For the past three weeks, I've been interested in imp-spirals in the big core size. My first attempt was Skimp 127: ;redcode-94x ;name Skimp 127 ;author Jay Han ;strategy A simple stone+imp t equ 12223 n equ 127 org s s spl.b 1 spl.b 1 spl.b 1 spl.b 1 spl.b 1 spl.b 1 mov.i -1, 0 spl.b 1 spl.b 2 jmp.b imp add.a #t, -1 d dat.f 0, 0 imp mov.i 0, t end As you may have noticed, this produces a 254-process, 127-point spiral, and nothing else (no stone, as advertised). I actually meant to make it a stone-spiral, but I mailed in the wrong file! I also meant to make it a single-layer imp-ring, but there was an extra SPL 1. Anyway, this program has been riding the Hill around 7th place for a while. My first submission, chronologically, was Sunburst 33, an overhauled warrior from the 88 Hill. I just modified the constants to fit the bigger core size. It's a simple stone with a multi-pass core-clear (one pass SPL0, then DAT-clear indefinitely). I had made major efforst to make it as small as it is, but such an effort may be wasted in the bigger core. ;redcode-94x ;name Sunburst 33 ;author Jay Han ;kill Sunburst p1 equ 27790 s1 equ 68 p2 equ 6914 s2 equ 55423 loop add inc, confuse confuse mov i1, >o1 mov.i >i1, >o1 mov.i >i1, >o1 mov.i >i1, >o1 mov.i >i1, >o1 mov.i front, front+9240+27720 mov.i back, back+9240+27720 spl.b front+9240 mov.i >i2, >o2 mov.i >i2, >o2 mov.i >i2, >o2 mov.i >i2, >o2 mov.i >i2, >o2 mov.i front, front-9240+27720 mov.i back, back-9240+27720 spl.b front-9240 mov.i >i3, >o3 mov.i >i3, >o3 mov.i >i3, >o3 mov.i >i3, >o3 mov.i >sp, >o3 mov.i >sp, >o3 mov.i >sp, >o3 sp jmp.b front+27720, e front dat 0,0 back dat 0,1 i1 dat 0, d i2 dat 0, d i3 dat 0, d o1 dat 0, front+9240 o2 dat 0, front-9240 o3 dat 0, front+27720 d mov.i 3, <27720 mov.i 2, >27720 djn.b -2, #9200 spl.b 0, <-5 dat <-6, <-7 e mov.i 2, <-2 jmp.b -1, <-7 dat <-1, <-1 end Next comes the Stimp series. The latest version, toward the botton of the Hill, is simply a stone/imp combination. The constants are adjusted so that the stone overwrites itself at one point, then checks to see if the imp is still alive; if it is, then the stone self-destructs; if not, it starts a new imp-spiral. Sorry for the bad formatting. The obfuscation at "hop" is because the stone has several processes in it, and I only want to start one single-threaded imp-ring, so the first process to execute "is" will make all the others die. ;redcode-94x ;name Stimp 2 ;author Jay Han ;kill Stimp ;strategy More Stone+Imp -> More Imps org start t equ 42547 st equ 34117 ; second imp-spiral x equ -2 ; Stone ends, imp is here boot equ 0 p2 equ 37332 q2 equ 37327 p1 equ 42816 q1 equ 7251 comp cmp.i -x ,-x-t stimp mov.i 0 ,st kill mov.i 2-pass (SPL0, DAT) coreclear org s0 bomb dat <-7 ,<-5 wait djn.b 0 ,#937 mov.i bomb ,3755 spl 0 ,<-5 mov.i @ptr ,=- e-mail: jayhan@cs.washington.edu Dept. of Computer Science and Engineering, FR-35. Fax: (206)543-2969 U. of Washington. Seattle, WA 98195, USA. [Sieg 230] (206)685-1224 Personal: 4820 Burke Ave N. Seattle, WA 98103, USA. (206)547-8559