From news-rocq.inria.fr!jussieu.fr!math.ohio-state.edu!howland.reston.ans.net!Germany.EU.net!main.Germany.EU.net!fu-berlin.de!news.th-darmstadt.de!uni-erlangen.de!kue!lrz-muenchen.de!news Fri Jul 19 10:33:02 1996 Article: 5632 of rec.games.corewar Path: news-rocq.inria.fr!jussieu.fr!math.ohio-state.edu!howland.reston.ans.net!Germany.EU.net!main.Germany.EU.net!fu-berlin.de!news.th-darmstadt.de!uni-erlangen.de!kue!lrz-muenchen.de!news From: guenzel@extern.lrz-muenchen.de (Bjoern Guenzel) Newsgroups: rec.games.corewar Subject: d-clear '2', only for the archive... Date: Thu, 18 Jul 1996 00:04:19 GMT Organization: Tichy Works Lines: 54 Distribution: world Message-ID: <4sjr8f$o5q@sparcserver.lrz-muenchen.de> NNTP-Posting-Host: dial048.ppp.lrz-muenchen.de X-Newsreader: Forte Free Agent 1.0.82 This is only for Planars archive, the principle is exactly the same as d-clear. It only has a better shape against bombers (at least I think so). THerefore it should score a bit better in the benchmarks, and that's the only reason I post it... Hm, perhaps it could do a decrement-protection? I wonder if it would be better to have no '>' in the spl? Perhaps I should do some more tests... Bjoern ;redcode-94 ;name d-clear II ;author bjoern guenzel ;strategy d-clear ;assert CORESIZE==8000 ;release 30.5.96 ;kill d-clear shift equ -3000 ptr dat 0,#-shift+1003 dat 0,0 dat 0,0 spl #0,>ptr loop mov.i bomb,>ptr djn.f loop,>ptr dat 0,0 dat 0,0 bomb dat >1,>bomb-ptr+2 bsource dat 0,-3 boot mov.i }ptr,>booptr mov.i }ptr,>booptr for 6 mov.i {bsource,{booptr rof booptr spl bsource+shift,ptr+shift mov.f #0,-1 dat >2000,>2500 z for (MAXLENGTH-CURLINE)/2 spl #z*100,#z*101 stp.a >0,#1 rof end boot