From ee31io@ee.surrey.ac.uk Thu Sep 19 12:24:18 1996 Received: from nez-perce.inria.fr (nez-perce.inria.fr [192.93.2.78]) by tobago.inria.fr (8.6.10/8.6.6) with ESMTP id MAA09548 for ; Thu, 19 Sep 1996 12:24:18 +0200 Received: from ainur.ee.surrey.ac.uk (ainur.ee.surrey.ac.uk [131.227.50.25]) by nez-perce.inria.fr (8.7.1/8.7.1) with SMTP id MAA03563 for ; Thu, 19 Sep 1996 12:24:07 +0200 (MET DST) Received: from duck10.ee.surrey.ac.uk by ainur.ee.surrey.ac.uk with smtp (Smail3.1.29.1 #4) id m0v3gGr-000OaYC; Thu, 19 Sep 96 11:23 BST Received: by duck10.ee.surrey.ac.uk (Smail3.1.29.1 #4) id m0v3gGq-0001gzC; Thu, 19 Sep 96 11:23 BST Message-Id: From: ee31io@ee.surrey.ac.uk (I M Oversby) Subject: Falcon v0.3 To: Damien.Doligez@inria.fr Date: Thu, 19 Sep 1996 11:23:47 +0100 (BST) X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Status: R Hi Planar, I posted this to the newsgroup a few days ago but I am not sure if it has appeared yet: ----------------------------------------------------------- I am a bit tired of seeing Falcon on the hill so I thought I should publish the code. The pcode is incredibly fragile so it should be easy to beat. The philosophy is simple; fight with a one-shot until it loses about 8 times more than it has won and then "permanently" switch to an imp/stone. I published earlier versions of both the one-shot and the imp/stone. The one-shot, Tsunami, now skips single bombs and has a 66% forward clear with 33% DJN-stream as opposed to the 50/50 of before. The gate is also a little better against die-hard imps I hope. I don't think it always does two SPL passes. I'm very sloppy when it comes to redcoding and testing! The imp/stone now has a mod-4 step to combat Blur, Scan Man and other mod-5 scanners against which I was losing really badly. As usual, this trades off against mod-4/mod-8 scanners and, I think, bombers and stones as the bombing phase takes longer. If you are wondering what the new anti-brainwash is, I removed the switch on being washed with 0. I now default to Tsunami when washed with 0. This helped against Marsden's brainwasher, Test, and also Barrage. I also lowered the mod-number to increase the chance of switching back to Tsunami when brainwashed with random numbers. I changed the boot distances and the p-code location but kept the pspace mod-numbers the same; I'm feeling generous (and lazy). Thats it. Enjoy, Ian ;redcode-94 quiet ;name Falcon v0.3 ;author Ian Oversby ;strategy v0.1 P-Warrior with two components ;strategy One-shot and imp / stone ;strategy v0.3 Trying new anti-brainwash ;assert 1 org str plc equ 3 gate1 equ (gate-3) tval equ 850 impstep equ 2667 sval equ 5600 ival equ 3000 dist equ 20 ;;------------------------------------------------------------ str LDP.A #plc, pval ;load wins/losses ADD.A pval, check MOD.A #3, check JMN.A boot, check ;permanent switch res LDP.AB #0, #0 sec SNE.AB #0, res ;check for loss lost ADD.A #3, pval SNE.AB #1, res ;check for win won SUB.A #3, pval STP.AB pval, #plc jump JMP.B tboot, <-100 pval DAT.F $0, $0 check DAT.F $7977, $0 ; 3 / 8 ;;--------------------- Tsunami , 25 lines ------------------------ for 10 DAT.F $0, $0 rof tboot MOV.I {tptr, b1 scan SNE.I *b1, @b1 DJN.F sloop, <-800 step SPL.B #-18, <-18 cloop MOV.I @bptr, >gate1 MOV.I @bptr, >gate1 bptr DJN.B cloop, {b1 ;;--------- Simple v0.2b with 16 process, 3 point spiral ---------- for 11 DAT.F $0, $0 rof iboot MOV.I {iptr,