From news-rocq!jussieu.fr!fu-berlin.de!newsfeed.tli.de!news.maxwell.syr.edu!howland.erols.net!portc02.blue.aol.com!audrey01.news.aol.com!not-for-mail Thu Jul 29 20:36:18 1999 Article: 10451 of rec.games.corewar Path: news-rocq!jussieu.fr!fu-berlin.de!newsfeed.tli.de!news.maxwell.syr.edu!howland.erols.net!portc02.blue.aol.com!audrey01.news.aol.com!not-for-mail From: happycan@aol.com (Happy can) Newsgroups: rec.games.corewar Subject: criticism/help please... Lines: 35 NNTP-Posting-Host: ladder05.news.aol.com X-Admin: news@aol.com Date: 26 Jul 1999 14:00:30 GMT Organization: AOL http://www.aol.com Message-ID: <19990726100030.20325.00001018@ng-fg1.aol.com> Xref: news-rocq rec.games.corewar:10451 ive given up on the replication chain idea...it is too slow and doesnt do enough to damage the other warrior, and self repair doesnt help much if one or two of the links are executing 8 SPL 0's in a row =P ive bene working on a paper/bomber, i have some code, but it doesnt quite work yet, i still need to get one MOV to point at the right address... The warrior moves itself 100 spaces down the core, then bombs the 24 addresses after itself with DATs, then repeats... the code: ;name confeti ;author marshall SPL 1 SPL 1 ;make 8 processes SPL 1 MOV }6, >6 ;replication MOV 5, >5 MOV 4, >4 ;carpet bombing MOV 3, >3 MOV 3, ??? ;this should point at the copy of the dat that has been incremented in bombing and replication JMP 101 ; this takes those 8 processes to the new copy DAT -6, 100 ;this dat gets incremented DAT -6, 100 ;this dat should be moved to the copy to replace the incremented copy this bit of code will replicate once, and could replicate infinitely if i could get that MOV to aim correctly. Would this warrior have any success? should it move farther down the core each time? is it highly innefficient? is it even written legally? any comments / critism is welcome =) Marshall From news-rocq!jussieu.fr!oleane!freenix!remarQ-easT!remarQ73!supernews.com!remarQ.com!remarQ69!WReNclone!WReNphoon3.POSTED!WReN!not-for-mail Thu Jul 29 20:36:21 1999 Article: 10457 of rec.games.corewar Path: news-rocq!jussieu.fr!oleane!freenix!remarQ-easT!remarQ73!supernews.com!remarQ.com!remarQ69!WReNclone!WReNphoon3.POSTED!WReN!not-for-mail From: Ben Ford Newsgroups: rec.games.corewar Subject: Re: criticism/help please... Organization: http://www.remarq.com: The World's Usenet/Discussions Start Here X-Originating-Host: 207.170.76.162 X-Wren-Trace: cOnO7vvptc31p73L+aWosq26qL+2vfGlubamu7mkqem0v/6mtv6sqbK+o72x+7r45/fh5P2W1u788Kjp9uf+7uE= Message-ID: <933019733.17353@www.remarq.com> References: <19990726100030.20325.00001018@ng-fg1.aol.com> Lines: 79 Date: Mon, 26 Jul 1999 12:08:49 -0800 NNTP-Posting-Host: 10.0.3.195 X-Complaints-To: wrenabuse@remarq.com X-Trace: WReNphoon3 933026519 10.0.3.195 (Mon, 26 Jul 1999 15:01:59 PDT) NNTP-Posting-Date: Mon, 26 Jul 1999 15:01:59 PDT Xref: news-rocq rec.games.corewar:10457 Marshal wrote: > ive given up on the replication chain idea...it is too > slow and doesnt do enough to damage the other warrior, > and self repair doesnt help much if one or two of the > links are executing 8 SPL 0's in a row =P Don't worry, we have all had brilliant ideas for warriors that end up failing miserably. At least you learned something from it. > ive bene working on a paper/bomber, i have some code, but > it doesnt quite work yet, i still need to get one MOV to > point at the right address... The warrior moves itself > 100 spaces down the core, then bombs the 24 addresses > after itself with DATs, then repeats... > > the code: > name confeti > author marshall > > SPL 1 > SPL 1 ;make 8 processes > SPL 1 > MOV }6, >6 ;replication > MOV 5, >5 > MOV 4, >4 ;carpet bombing > MOV 3, >3 > MOV 3, ??? ;this should point at the copy of the dat that > has been incremented in bombing and replication > JMP 101 ; this takes those 8 processes to the new copy > DAT -6, 100 ;this dat gets incremented > DAT -6, 100 ;this dat should be moved to the copy to > replace the incremented copy > > this bit of code will replicate once, and could replicate > infinitely if i could get that MOV to aim correctly. > Would this warrior have any success? should it move > farther down the core each time? is it highly > innefficient? is it even written legally? any comments > / critism is welcome =) > > Marshall It could have success. It probably won't score very high on its own since hill papers generally need to have q-scans to get on, but it should be able to make the beginner hill once its tuned. You should try out different step sizes to see what one is best, it probably will work out better as being bigger since most other papers have larger steps than that. It is highly inefficient. It seems to be legal. Here is what I would do. SPL 1 SPL 1 ;make 8 processes SPL 1 p1 SPL @0, >100 ; step size and weak increment attack MOV }p1, >p1 ;replication MOV pb, >p1 MOV pb, >p1 ;carpet bombing MOV pb, >p1 MOV {p1, mov style of splitting to the new copies is a lot faster than having to reset copy distances. Since the spl is copied first and is done with post-increment, the copied spl will still be @0, >100. -Ben * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network * The fastest and easiest way to search and participate in Usenet - Free!