> ;name Chainlock v02a
Hey, a replicated imp/stone. I like the mix of strategies, and also the
way
this makes efficient use of the Silk parallel processes. Once it gets
going
the stepped bombing looks strong.
> Plch: spl 1, <qb1
> qtab2 spl 1, <qb2
> spl 1, <qb3
8 processes.
> spl @[EMAIL PROTECTED]
> mov }-1,>-1
Standard Silk front end creating a strip of replicators, each offspring
having 8 processes and executing 1 step behind its parent. The first
replicator may be vulnerable because it is sitting next to the
quickscan; it might be worth putting in a gap here or booting (which would
add only 16 cycles and 2 words).
Given the process-hungry SPL #0 in the payload the silk stage won't have
time to create all that many Silk/Imps, but enough to crush DAT bombers
and
start imp rings in several places.
> head: spl #-1143,<dec
> mov bmb,@[EMAIL PROTECTED]
this with 8 processes only achieves 1 decrement and 1 bomb in
the
first 24 cycles! Coupled with the Silk engine making 4 copies in parallel
it means about a hundred cycles before the stepped bombing starts, even
longer until the first imps are launched. That gives a scanner quite a
long
time to score a stun and I think its the most significant weakness.
> sub.f head,igo
Step the bombing (because the Silk launch means SUB and MOV execute
alternately). This is great once it gets going, and a neat trick I've
seen
to speed this up is to step your offspring's bombing not your own, with
something like
sub.f head, igo+dest1
This means that all papers after the first step every bomb rather than
wasting the first 8 bombs on top of each other. I think you could also
swap
the order of the MOV and SUB, and use
sub.f head, igo-dest1
to step your parent (though I've no idea if this has any advantages in
practice).
Both these approaches lose some bombing efficiency once the papers start
getting stunned, but I suspect the advantage of a faster start to bombing
will more than pay for this -- if a scanner hasn't achieved a solid stun
within a few hundred cycles the bombing is likely to kill it.
Something else you might try is
mov bmb @[EMAIL PROTECTED]
would fall back to strip bombing if the stepping is damaged.
> igo: djn.f imp-(8*1143),{trail
> bmb: dat {bstep,<1
> imp: mov.i #1,1143
Launch 7pt imp rings. This probably makes sense given you have 8
processes
available and that 3pt are more common, though they don't form a proper
spiral. Do your constants make some of the rings overlap? This could
strengthen them.
If you want to try a more fundamental change you could slim down to 5
processes and use
spl @[EMAIL PROTECTED]
mov }-1,>-1
mov }-2,>-2
to launch -- it might even be possible to use 4 processes but you would
have
to do without the bomb. Either of these launches would get everything
moving quite a lot faster, though they restrict imps to 3pt and need a
whole
new set of constants.
Lots of potential 8-)
Robert
----- Original Message -----
From: <neogryzor@[EMAIL PROTECTED]
>
To: "Multiple recipients of list COREWAR-L"
<corewar-l@[EMAIL PROTECTED]
>
Sent: Saturday, March 29, 2008 1:59 AM
Subject: Chainlock v02a, a strong paper/imp for '94 hills
> Just to post something and promote interest on CW.
> This Paper/imp was pushed off the 94'nop hill some time ago but i
> think it is still a hard opponent.
>
> Also found a file with its results when it entered the hill, it may be
> interesting.
>
> If you still have those old paper/imps in your benchmark, kick them
> and put this one instead, it is more effective. ;)
>
> Enjoy
>
> ;redcode-94nop
> ;name Chainlock v02a
> ;author G.Labarga
> ;strategy Q4.5->paper/imp
> ;assert CORESIZE==8000
> ;strategy 1. First test version
> ;strategy 2. well optimized
>
> zero equ qbomb
> qtab3 equ qbomb
>
> qbomb dat >qoff , >qc2
> dat 0,0
> dat 0,0
>
> dest1 equ 6065
> ;dest2 equ 4450
> dec equ 5905
> trail equ 6876
> bstep equ 5353 ;-6
>
> Plch: spl 1, <qb1
> qtab2 spl 1, <qb2
> spl 1, <qb3
>
> spl @[EMAIL PROTECTED]
> mov }-1,>-1
> head: spl #-1143,<dec
> mov bmb,@[EMAIL PROTECTED]
> sub.f head,igo
> igo: djn.f imp-(8*1143),{trail
> bmb: dat {bstep,<1
> imp: mov.i #1,1143
>
> for 4
> dat 0 , 0
> rof
> dat zero-1 , qa1
> qtab1 dat zero-1 , qa2
>
> for 17
> dat 0 , 0
> rof
> for 14
> dat 0 , 0
> rof
> for 14
> dat 0 , 0
> rof
>
> ;------- Q^4.5 ------
> qc2 equ ((1+(qtab3-qptr)*qy)%CORESIZE)
> qb1 equ ((1+(qtab2-1-qptr)*qy)%CORESIZE)
> qb2 equ ((1+(qtab2-qptr)*qy)%CORESIZE)
> qb3 equ ((1+(qtab2+1-qptr)*qy)%CORESIZE)
> qa1 equ ((1+(qtab1-1-qptr)*qy)%CORESIZE)
> qa2 equ ((1+(qtab1-qptr)*qy)%CORESIZE)
> qz equ 2108
> qy equ 243 ;qy*(qz-1)=1
>
> ;q0 mutation
> qgo sne qptr+qz*qc2 , qptr+qz*qc2+qb2
> seq <qtab3 , qptr+qz*(qc2-1)+qb2
> jmp q0 , }q0
> sne qptr+qz*qa2 , qptr+qz*qa2+qb2
> seq <qtab1 , qptr+qz*(qa2-1)+qb2
> jmp q0 , {q0
> sne qptr+qz*qa1 , qptr+qz*qa1+qb2
> seq <(qtab1-1) , qptr+qz*(qa1-1)+qb2
> djn.a q0 , {q0
> ;q1 mutation
> sne qptr+qz*qb3 , qptr+qz*qb3+qb3
> seq <(qtab2+1) , qptr+qz*(qb3-1)+(qb3-1)
> jmp q0 , }q1
> sne qptr+qz*qb1 , qptr+qz*qb1+qb1
> seq <(qtab2-1) , qptr+qz*(qb1-1)+(qb1-1)
> jmp q0 , {q1
>
> sne qptr+qz*qb2 , qptr+qz*qb2+qb2
> seq <qtab2 , qptr+qz*(qb2-1)+(qb2-1)
> jmp q0
> ;qz mutation
> seq >qptr , qptr+qz+(qb2-1)
> jmp q2 , <qptr
> ;q0 mutation
> seq qptr+(qz+1)*(qc2-1),qptr+(qz+1)*(qc2-1)+(qb2-1)
> jmp q0 , }q0
> seq qptr+(qz+1)*(qa2-1),qptr+(qz+1)*(qa2-1)+(qb2-1)
> jmp q0 , {q0
> seq qptr+(qz+1)*(qa1-1),qptr+(qz+1)*(qa1-1)+(qb2-1)
> djn.a q0 , {q0
> jmz.f Plch , qptr+(qz+1)*(qb2-1)+(qb2-1)
>
> qoff equ -86
> qstep equ -7
> qtime equ 19
>
> q0 mul.b *2 , qptr
> q2 sne {qtab1 , @[EMAIL PROTECTED]
> q1 add.b qtab2 , qptr
> mov qtab3 , @[EMAIL PROTECTED]
> qptr mov qbomb , }qz
> sub #qstep , qptr
> djn -3 , #qtime
> djn.f Plch , #0
>
> end qgo
>
> And the scores:
> -----------------------
>
> Program "Chainlock v02a" (length 100) by "G.Labarga"
> (contact address "neogryzormail@[EMAIL PROTECTED]
"):
> has challenged the ICWS '94 No Pspace hill.
>
> Program "Hullabaloo" (length 100) by "Roy van Rijn"
> (contact address "roy_van_rijn@[EMAIL PROTECTED]
"):
> ;strategy 00000075-154.05
> Chainlock v02a wins: 54
> Hullabaloo wins: 36
> Ties: 160
>
> Program "The Humanizer" (length 100) by "bvowk/fiz"
> (contact address "fizmo@[EMAIL PROTECTED]
"):
> ;strategy partly evolved
> Chainlock v02a wins: 76
> The Humanizer wins: 24
> Ties: 150
>
> Program "Borgir" (length 100) by "Christian Schmidt"
> (contact address "fizmo_master@[EMAIL PROTECTED]
"):
> ;strategy
> Chainlock v02a wins: 56
> Borgir wins: 26
> Ties: 168
>
> Program "New-T" (length 100) by "Roy van Rijn"
> (contact address "roy_van_rijn@[EMAIL PROTECTED]
"):
> ;strategy Imp/Stone
> ;strategy Based on the good old Newt
> Chainlock v02a wins: 19
> New-T wins: 35
> Ties: 196
>
> Program "S.D.N. " (length 100) by "Christian Schmidt "
> (contact address "fizmo@[EMAIL PROTECTED]
"):
> ;strategy "Satellite Defending Network"
> ;strategy a revolutionary strategical concept
> ;strategy which is much faster and less visible
> ;strategy I am still in progress to discover
> ;strategy this new and interesting field of
> ;strategy 'satellite using' warriors
> ;strategy including wipe check of the satellites
> ;strategy optimized test warrior
> Chainlock v02a wins: 23
> S.D.N. wins: 25
> Ties: 202
>
> Program "Arrow" (length 100) by "Christian Schmidt"
> (contact address "fizmo_master@[EMAIL PROTECTED]
"):
> ;strategy scanner using John Metcalf's Zooom-trick
> Chainlock v02a wins: 84
> Arrow wins: 80
> Ties: 86
>
> Program "Riccia" (length 100) by "Zul Nadzri"
> (contact address "sayembara@[EMAIL PROTECTED]
"):
> ;strategy November 2004
> Chainlock v02a wins: 123
> Riccia wins: 43
> Ties: 84
>
> Program "Gargantuan" (length 100) by "Roy van Rijn"
> (contact address "roy.van.rijn@[EMAIL PROTECTED]
"):
> ;strategy qScan into impPaper & coreclearingPaper & stone
> Chainlock v02a wins: 56
> Gargantuan wins: 28
> Ties: 166
>
> Program "Cyborgs Rule" (length 100) by "bvowk/Roy"
> (contact address "roy_van_rijn@[EMAIL PROTECTED]
"):
> ;strategy Partly human, partly evolved
> ;strategy I want to try it too, intresting paper
> ;strategy Substracted the paper, made a new boot and added a qscan
> ;strategy Optimized constants
> Chainlock v02a wins: 77
> Cyborgs Rule wins: 27
> Ties: 146
>
> Program "Maelstrom" (length 100) by "Roy van Rijn"
> (contact address "roy_van_rijn@[EMAIL PROTECTED]
"):
> ;strategy qScan into impPaper & coreclearingPaper
> Chainlock v02a wins: 25
> Maelstrom wins: 28
> Ties: 197
>
> Program "again" (length 100) by "Lukasz Grabun"
> (contact address "lgrabun@[EMAIL PROTECTED]
"):
> Chainlock v02a wins: 25
> again wins: 17
> Ties: 208
>
> Program "Black Knight" (length 100) by "Christian Schmidt"
> (contact address "fizmo_master@[EMAIL PROTECTED]
"):
> ;strategy qscanner
> ;strategy Carbonite-styled stone
> ;strategy launching a-imps AND b-imps
> Chainlock v02a wins: 35
> Black Knight wins: 28
> Ties: 187
>
> Program "paper(paper(paper(clear)))" (length 100) by "Sascha Zapf"
> (contact address "nc-zapfsa@[EMAIL PROTECTED]
"):
> ;strategy Q4.5 -> Coreclearing Paper's
> Chainlock v02a wins: 26
> paper(paper(paper(clear))) wins: 25
> Ties: 199
>
> Program "thoughts of fire" (length 100) by "John Metcalf"
> (contact address "grumpy3039@[EMAIL PROTECTED]
"):
> ;strategy scanner
> Chainlock v02a wins: 83
> thoughts of fire wins: 45
> Ties: 122
>
> Program "Hydra II ( Test )" (length 100) by "Sascha Zapf"
> (contact address "nc-zapfsa@[EMAIL PROTECTED]
"):
> ;strategy Paper/Stone
> Chainlock v02a wins: 60
> Hydra II ( Test ) wins: 31
> Ties: 159
>
> Program "HazyLazy C 11" (length 100) by "CS SG"
> (contact address "steveg58@[EMAIL PROTECTED]
"):
> ;strategy Mod 10 cmp scanner shifting to D-Clear
> ;strategy Q4(?) scan an constants by Christian
> Chainlock v02a wins: 83
> HazyLazy C 11 wins: 50
> Ties: 117
>
> Program "Surfing on the net" (length 100) by "Fizmo&Neo"
> (contact address "neogryzormail@[EMAIL PROTECTED]
"):
> ;strategy S.D.N. system
> Chainlock v02a wins: 51
> Surfing on the net wins: 29
> Ties: 170
>
> Program "Spiker" (length 100) by "Roy van Rijn"
> (contact address "roy_van_rijn@[EMAIL PROTECTED]
"):
> ;strategy Q4.5/Stunbomber
> ;strategy Lets try something different from papers
> ;strategy And StrikeTwo is pushed off, so lets try a bomber!
> ;strategy Formed after Behemot/RoVanquisher
> Chainlock v02a wins: 96
> Spiker wins: 46
> Ties: 108
>
> Program "Pain A Test 25" (length 95) by "Steve Gunnell"
> (contact address "steveg58@[EMAIL PROTECTED]
"):
> ;strategy Agony style CMP scanner with CW 84 style Q4 scan.
> Chainlock v02a wins: 118
> Pain A Test 25 wins: 46
> Ties: 86
>
> Program "soft-grain" (length 100) by "John Metcalf"
> (contact address "grumpy3039@[EMAIL PROTECTED]
"):
> ;strategy scanner
> Chainlock v02a wins: 89
> soft-grain wins: 37
> Ties: 124
>
> Program "Chainlock v02a" (length 100) by "G.Labarga"
> (contact address "neogryzormail@[EMAIL PROTECTED]
"):
> ;strategy Qscan->paper/imp
> ;strategy First test version
> Chainlock v02a wins: 28
> Ties: 196
>
> Your overall score: 135.085714
>
> soft-grain has been pushed off the ICWS '94 No Pspace hill.
> The current ICWS '94 No Pspace hill:
> Last battle concluded at : Thu Nov 11 11:25:46 EST 2004
>
> # %W/ %L/ %T Name Author Score Age
> 1 35/ 27/ 37 Hullabaloo Roy van Rijn 143 118
> 2 33/ 27/ 40 The Humanizer bvowk/fiz 140 6
> 3 34/ 29/ 37 Borgir Christian Schmidt 138 153
> 4 25/ 14/ 62 Chainlock v02a G.Labarga 135 1
> 5 30/ 25/ 45 New-T Roy van Rijn 135 52
> 6 41/ 47/ 13 Arrow Christian Schmidt 134 321
> 7 25/ 17/ 58 S.D.N. Christian Schmidt 134 43
> 8 30/ 28/ 42 Gargantuan Roy van Rijn 132 139
> 9 38/ 44/ 18 Riccia Zul Nadzri 132 10
> 10 24/ 16/ 59 Maelstrom Roy van Rijn 132 262
> 11 30/ 30/ 40 Cyborgs Rule bvowk/Roy 131 31
> 12 28/ 26/ 46 again Lukasz Grabun 130 11
> 13 25/ 20/ 55 Black Knight Christian Schmidt 130 143
> 14 26/ 23/ 51 paper(paper(paper(clear)) Sascha Zapf 129 222
> 15 37/ 44/ 19 thoughts of fire John Metcalf 129 18
> 16 37/ 46/ 17 HazyLazy C 11 CS SG 129 330
> 17 30/ 32/ 38 Hydra II ( Test ) Sascha Zapf 129 51
> 18 30/ 31/ 39 Surfing on the net Fizmo&Neo 128 3
> 19 36/ 45/ 19 Spiker Roy van Rijn 127 55
> 20 36/ 48/ 16 Pain A Test 25 Steve Gunnell 124 2
>
> 21 33/ 44/ 23 soft-grain John Metcalf 123 20
>
> Not bad for such a defensive strategy.
>
>
|
