Sections
Home
Hills
Infinite Hills
Tournaments
Software
Evolving
Optimizer
Community
Newsletter
Discussion
History
Sections
 
For Beginners
First Steps
FAQ
Guides
Lexicon
Benchmarks
For Beginners
> Home > The Corewar Newsletters > Core Warrior > Issue #1

Issue 44                                                    September 23, 1996
______________________________________________________________________________
Core Warrior is a weekly newsletter promoting the game of corewar.  Emphasis
is placed on the most active hills--currently the '94 draft hill and the
beginner hill.  Coverage will follow where ever the action is.  If you have
no clue what I'm talking about then check out these five-star internet locals
for more information:

FAQs are available by anonymous FTP from rtfm.mit.edu as
pub/usenet/news.answers/games/corewar-faq.Z
FTP site is: ftp.csua.berkeley.edu /pub/corewar
Web pages are at: (Please note new Stormking's address)
http://www.koth.org/                            ;Stormking
http://www.ecst.csuchico.edu/~pizza/koth        ;Pizza
http://pauillac.inria.fr/~doligez/corewar/      ;Planar

Newbies should check the stormking page for the FAQ, language specification,
guides, and tutorials.  Post questions to rec.games.corewar.  All new players
are infinitely welcome!

If ftp.csua.berkeley.edu is unreachable, you can download pMARS at:
Terry's web page--http://www.infi.net/~wtnewton/corewar/
Planar ftp site--ftp://ftp.inria.fr/INRIA/Projects/para/doligez/cw/pmars
Fechter ftp site--ftp://members.aol.com/ofechner/corewar

A collection of Bezzi's hints in the first issues is available at:
ftp://ftp.volftp.vol.it/pub/pc/msdos/games/solutions/bbhints.zip
______________________________________________________________________________
Greetings.

After a forced break in late summer, due to Pizza maintenance, the 94 hill
has gone back to his standard level of activity; don't be fouled by the
small number of new warriors, more than 50 test that would have made the
hill have been submitted. The test option proved in fat very popular even if
there are a few drawbacks; one is that the aging of warriors cannot be
ompared with that of past ones being now far more difficult. The second is
that opponent don't have chanes to challenge a warrior under test, as Paul
Kline pointed out, until it's released in offiial form. Last, as I
experimented, a test warrior can score very different once it's released,
this because test works on but 100 fights and the real fight on 200, and I'm
not very lucky having lost more than 10 points :-(

Steven Morrell volunteered to recalculate the hall of fame removing
unnatural aging from the scores one we agree on a method to do it. The
better one seems to disount aging if the pushed off warrior has an score
under a treshold; I proposed 125 but Planar told me that some warriors,
Impfinity to tell one, have been pushed off with lower scores. Having to set
one, to allow Steven start his work, what about 120.

Another hot topic in the newsgrou is if there is a need of a central
organization, to take the place of the near dead ICWS, or if we can go on
without it. The discussion is still hot.

For those loving the old 88 redcode standard; it's worth noting that the 88
hill, available at koth@koth.org (Stormking), has a new leader with an
incredible 17 pts lead. Congratulations to David Moore for his Racer X.


--Beppe Bezzi
______________________________________________________________________________
Current Status of the Internet Pizza Server ICWS '94 Draft Hill:

Hill Specs:
   coresize: 8000
   max. processes: 8000
   duration: after 80,000 cycles, a tie is declared.
   max. entry length: 100
   minimum distance: 100
   rounds fought: 200
   instruction set: ICWS '94 Draft

The current ICWS '94 Draft hill:
 #   %W /  %L /  %T                       Name               Author Score  Age
 1  40.3/ 19.0/ 40.7                    Gigolo   Core Warrior staff 161.6    1
 2  38.8/ 20.2/ 41.0           unrequited love                kafka 157.3   45
 3  42.8/ 32.0/ 25.2                  CC Paper                Franz 153.5    4
 4  42.1/ 32.0/ 25.9                     Frogz                Franz 152.2    5
 5  31.5/ 14.8/ 53.7      Return Of The Jedimp             John K W 148.2  113
 6  44.1/ 40.6/ 15.3                    Blur 2        Anton Marsden 147.6  124
 7  42.8/ 40.2/ 17.0                 Goldfinch              P.Kline 145.3   66
 8  42.4/ 39.4/ 18.2                     Probe        Anton Marsden 145.3  125
 9  31.5/ 19.7/ 48.8                   Rosebud                Beppe 143.2  798
10  35.9/ 28.6/ 35.6              Simple v0.4b          Ian Oversby 143.2   76
11  38.7/ 35.3/ 26.1              Flimsy v0.5c          Ian Oversby 142.1    6
12  43.4/ 47.2/  9.4                  Q^2 Miro         Anders Ivner 139.6  175
13  30.0/ 20.5/ 49.6               Impish v0.1          Ian Oversby 139.4    2
14  39.3/ 39.3/ 21.4                  mrb-test           m r bremer 139.4   31
15  39.1/ 40.2/ 20.7                      Test        Anton Marsden 138.1   35
16  25.1/ 12.6/ 62.3                    ompega       Steven Morrell 137.6  240
17  33.8/ 30.3/ 35.9        Jack in the box II          Beppe Bezzi 137.2  482
18  33.8/ 30.7/ 35.5               Falcon v0.3          Ian Oversby 136.9   30
19  38.3/ 40.3/ 21.4          Instant Wolf 3.4                Edgar 136.4    8
20  37.8/ 39.8/ 22.4               Inferno 2.4       Philip Kendall 135.9    3
21  36.1/ 39.0/ 24.9                T.N.T. pro    Maurizio Vittuari 133.2  843
22  39.6/ 46.2/ 14.2           Earthquake v0.2         Bjoern & Ian 132.9   82
23  38.3/ 45.3/ 16.3                 myVamp5.5             Paulsson 131.3   15
24  35.7/ 40.7/ 23.6                 Yogi Bear              P.Kline 130.8  329
25  36.7/ 44.0/ 19.3          BackFromVacation                Edgar 129.5   17

Weekly age: 15 ( 9 last issue, 9 the month before ) more than 50 test
New warriors: 8  Turnover/age rate 53%
Average age:   146 ( 174 last issue, 180 the week before )
Average score: 142 ( 136 last issue, 142 the week before )
The top 25 warriors are represented by several authors :-) Sorry, I don't
have the patience to fill in the details this week.

Lots of new warriors in the top position. Bezzi made many test during the
week, some of which scored very high, to shodown with the new King, Gigolo,
a cooperative warrior made with the others of Core Warrior staff. Don't ask
him why after scoring 12 points over the second his warrior setted with but
three unless you want to learn some new Italian words :-)
Kafka's unrequited love loses top position after two weeks of domain, but
remains near the top. Franz puts two warriors in top ranks, both using the
qscan->silk combination; perhaps replicators are back vith vengeance.
Q^2 scans are dominating the hill at expense of pspacers.
______________________________________________________________________________
94 - What's New

 #   %W /  %L /  %T                       Name               Author Score  Age
 1  37.2/ 19.8/ 43.0                    Gigolo   Core Warrior staff 154.5    1
 2  42.8/ 31.4/ 25.8                     Frogz                Franz 154.3    1
 4  41.5/ 32.9/ 25.5                  CC Paper                Franz 150.1    1
 5  39.3/ 35.2/ 25.5              Flimsy v0.5c          Ian Oversby 143.3    1
 8  38.9/ 39.3/ 21.8          Instant Wolf 3.4                Edgar 138.5    1
13  27.4/ 21.5/ 51.1               Impish v0.1          Ian Oversby 133.3    1
17  36.6/ 41.0/ 22.4               Inferno 2.4       Philip Kendall 132.1    1
17  37.9/ 45.4/ 16.7                 myVamp5.5             Paulsson 130.5    1

Revolution in the first spots, we have a new 1st, 3rd and 4th.
______________________________________________________________________________
94 - What's No More

 #   %W /  %L /  %T                       Name               Author Score  Age
26  35.3/ 45.6/ 19.1                   Twister          Beppe Bezzi 125.0  575
26  23.6/ 21.1/ 55.3                 Armory II             John K W 126.0  257
26   1.5/  1.7/  0.8                 myVamp5.4             Paulsson   5.4  159
26   0.2/  1.0/  2.8                 September                Beppe   3.4   21
26  36.1/ 46.9/ 17.0       Violent Micro v0.4d             basehead 125.2   16
26   1.8/  1.8/  0.4          Instant Wolf 3.4                Edgar   5.9   15
26   1.6/  2.1/  0.3              Flimsy v0.5a          Ian Oversby   5.2   14
26   4.0/  2.5/  1.5            Atom Smasher 2        Anton Marsden  13.5    3

Twister leaves us at the age of 575 and Armory at 257; all other are very
young or killed by the author.
______________________________________________________________________________
94 - What's Old

 #   %W /  %L /  %T                       Name               Author Score  Age
21  36.1/ 39.0/ 24.9                T.N.T. pro    Maurizio Vittuari 133.2  843
 9  31.5/ 19.7/ 48.8                   Rosebud                Beppe 143.2  798
17  33.8/ 30.3/ 35.9        Jack in the box II          Beppe Bezzi 137.2  482
24  35.7/ 40.7/ 23.6                 Yogi Bear              P.Kline 130.8  329
16  25.1/ 12.6/ 62.3                    ompega       Steven Morrell 137.6  240

Rosebud keeps his position while T.N.T. pro and Yogi lose some spots and
Jack a few. Seen the sarcity of veterans I decided to include them starting
at 200 age.
______________________________________________________________________________
HALL OF FAME
* means the warrior is still active.

Pos    Name                  Author          Age     Strategy
 1  Thermite II            Robert Macrae     2262    Qscan -> bomber
 2  Impfinity v4g1         Planar            1993    Stone/ imp
 3  Jack in the box        Beppe Bezzi       1620    P-warrior
 4  Tornado 3.0            Beppe Bezzi       1567    Bomber
 5  Torch t18              P.Kline           1539    Bomber
 6  Chameleon              Myer R Bremer     1437    P-warrior
 7  Frontwards v2          Steven Morrell    1420    One shot scanner
 8  Evol Cap 6.6           John Wilkinson    1299    Imp / stone
 9  quiz                   Schitzo           1262    Scanner/ bomber
10  T.N.T.                 Maurizio Vittuari 1204    Bomber
11  Grilled Octopus v0.5   David Boeren      1154    P-warrior
12  Hazy Shade II          John Wilkinson    1102    P-warrior
13  Stepping Stone         Kurt Franke       1049    Qscan -> Vampire
14  Iron Gate 1.5          Wayne Sheppard     926    CMP scanner
15  Agony II               Stefan Strack      912    CMP scanner
16  Barrage                Anton Marsden      876    Qscan -> replicator
17  Blue Funk              Steven Morrell     869    Stone/ imp
18  T.N.T. pro             Maurizio Vittuari  843 *  Bomber
19  Flurry                 Anton Marsden      835    Qscan -> pwarrior
20  Thermite 1.0           Robert Macrae      802    Qscan -> bomber
21  Rosebud                Beppe Bezzi        798 *  Stone/ imp
22  Blue Funk 3            Steven Morrell     766    Stone/ imp
23  Night Train            Karl Lewin         755    Replicator
24  Mirage 1.5             Anton Marsden      736    Scanner/ bomber
25  Blizzard               Anton Marsden      713    Qscan -> replicator

T.N.T. pro gains a spot; no new entries and we won't have for long being
next scheduled at more than 200 distance.
______________________________________________________________________________
Current Status of the Internet Pizza Server Beginner's Hill:

Hill Specs:
   coresize: 8000
  max. processes: 8000
   duration: after 80,000 cycles, a tie is declared.
max. entry length: 100
 minimum distance: 100
    maximum age: At age 100, warriors are retired.
   rounds fought: 200
  instruction set: ICWS '94 Draft

The current Beginner hill:
 #   %W /  %L /  %T                       Name               Author Score  Age
 1  54.6/ 34.4/ 11.0                 Throw Rug             J E Long 174.8   30
 2  55.2/ 36.1/  8.7              Tsunami v0.3          Ian Oversby 174.3   73
 3  53.4/ 35.2/ 11.4       Violent Micro v0.4d             basehead 171.6   44
 4  48.9/ 33.3/ 17.8               Heretic 1.0       Philip Kendall 164.6   55
 5  44.8/ 31.8/ 23.3           Versatility 1.6                 Ross 157.8   42
 6  44.8/ 33.2/ 22.0               Inferno 2.3       Philip Kendall 156.5   31
 7  46.3/ 37.8/ 15.9          Velveeta Shift-I                 shar 154.7    6
 8  48.7/ 42.8/  8.5                  Hawk 1.0       Philip Kendall 154.7   27
 9  39.5/ 25.6/ 34.9                  Gravel 7           Justin Kao 153.5   50
10  44.7/ 38.8/ 16.6           BloodSucker 1.0       Philip Kendall 150.6   54
11  37.3/ 25.9/ 36.8    Ties, Ties, Ties! (+3)                 Ross 148.7   92
12  44.5/ 40.9/ 14.6   Neverending Brutality 7                Franz 148.1   21
13  42.3/ 36.7/ 21.0                Black v0.1           Justin Kao 147.9   45
14  37.6/ 28.2/ 34.1                 Papyrus 6           Justin Kao 147.0   61
15  41.9/ 38.7/ 19.4             Vampirism 1.4       Philip Kendall 145.0   48
16  41.1/ 37.3/ 21.5                       RSB             harleyQ2 144.9   15
17  44.6/ 46.3/  9.1               T-1 w/decoy                 Ross 143.0   83
18  43.9/ 46.4/  9.6                       T-1                 Ross 141.4   91
19  39.9/ 39.2/ 20.9                  (-:  :-)                 Ross 140.5   79
20  38.7/ 38.7/ 22.6               TIE Fighter                 Ross 138.8   78
21  33.9/ 35.5/ 30.6                      Frog                Franz 132.4    8
22  40.0/ 52.1/  8.0                Syzygy 1.2       Philip Kendall 127.9   29
23  32.2/ 38.5/ 29.4           Hammermill v2.4        Matt Lewinski 125.9    2
24  34.8/ 53.4/ 11.9                  jmztest2         George Eadon 116.1    3
25   4.0/  0.0/  0.0               Epithelioma        Matt Lewinski  12.0    1

Ick died of old age. The scanner Throw rug is the new leader
______________________________________________________________________________
The Hint
Killing imps
by Beppe Bezzi

Imp stones are proliferating in the 94 hill, sometimes with a qscan
sometimes not. Against the qscan the only defence is booting away as fast as
you can, little else to say. Against stones, apart being a replicator, one
have to be resistant against decrements/increments and djn streams, I'll
speak of that in future. Against imps the best defence is a good gate.

94 standard allows combining the gate with a forward running coreclear, but
there are a few different ways to do that, what's the best ?

I ran a few test pitting the more common gate clears against some different
imp spirals; the spirals are all 3 points, 8 processes, binary launched and
differ but in the imp instruction:

Spiral #0,2667          mov.i   #0,     2667 
Spiral #1,2667          mov.i   #1,     2667
Spiral #-1,2667         mov.i   #-1,    2667
Spiral #2667,*0         mov.i   #2667,  *0

One can think that the first three will behave in the same way, but that's
not true as we'll see. The last one is a different kind that's getting more
and more popular; it's the a-driven imp, used by Die Hard and now by Rosebud.

Now the clears. All those clears can easily be fitted inside your favourite
bomber or scanner.

;name   gate c-c1

gate1   equ     (gate-5)
	org     start
for 10
	dat 0,0
rof
	dat     -5,   100
gate    dat     -5,   jump-gate1+5
	spl     #-5,  jump-gate1+5 
for 10
	dat     0,0             ;the warrior goes here
rof
start   spl     #-10,#10
	mov     @jump,>gate1
jump    djn     start+1,{gate+1

That's the clear used by Torch 18 and Thermite 2, just to say two successful
warriors, it's a two pass spl/dat clear, switching to dat one its djn stream
derements the b-field of 'jump' line. 
Here is how it scored:

Rank    Name                    Author                   %W  %L  %T   Score
___________________________________________________________________________
  1     gate c-c1               Anonymous                85   1  14   1074  
  2     Spiral #2667,*0         Anonymous                 1  45  54   57    
  3     Spiral #0,2667          Anonymous                 1  98   1   4     
  4     Spiral #1,2667          Anonymous                 1  98   1   4     
  5     Spiral #-1,2667         Anonymous                 1  98   1   4     

You can see it's very effective against mov #xx,2667 imps, when the imp hits
the gate the  attack is transferred to the next branch of the spiral, but
scores less than 50% wins against a-driven imps.

;name   gate c-c2
;assert 1
gate1   equ     (gate-5)
	org     start
for 10
	dat 0,0
rof
	dat     -5,   100
gate    dat     -5,   jump-gate1+5
	spl     #-5,  jump-gate1+5 
for 10
	dat     0,0
rof
start   spl     #-10,#10
	mov     @jump,>gate1
	mov     @jump,>gate1
jump    djn     start+1,{gate+1

This clear is the same of gate c-c1 but has one more mov line. This makes
the clear more vulnerable against bombs, but improves our performances
against both silk and imps. Tornado 3.0 used this clear.


Rank    Name                    Author                   %W  %L  %T   Score
___________________________________________________________________________
  1     gate c-c2               Anonymous                92   1   8   1128  
  2     Spiral #2667,*0         Anonymous                 1  69  30   33    
  3     Spiral #0,2667          Anonymous                 1  99   0   3     
  4     Spiral #1,2667          Anonymous                 1  99   0   3     
  5     Spiral #-1,2667         Anonymous                 1  99   0   3     

As you can notice the a-driven imp has less chances to slip through the gate.

;name   gate c-ca
;assert 1
gate1   equ     (gate-5)
	org     start
for 10
	dat 0,0
rof
	dat     100
gate    dat     jump-gate1+5,-5
	spl     #jump-gate1+5,-5
for 10
	dat     0,0
rof
start   spl     #-10,#10
	mov     @jump,}gate1
jump    djn     start+1,<gate+1

This is the same clear of c-c1, but it's driven by the a-field of the gate
line. Frontwards used this clear.

Rank    Name                    Author                   %W  %L  %T   Score
___________________________________________________________________________
  1     gate c-ca               Anonymous                41   4  56   711   
  2     Spiral #-1,2667         Anonymous                 8  14  78   102   
  3     Spiral #1,2667          Anonymous                 5  18  77   92    
  4     Spiral #0,2667          Anonymous                 1  33  66   69    
  5     Spiral #2667,*0         Anonymous                 1  98   1   4     

It's very effective against a-driven imps but much less against classic
ones, expecially against mov.i #-1,2667 because the gate copies an imp
instruction to attack the imp itself, and this has, obviously, no effect.

;name stargate
;assert 1

org stun

gate    dat     stun,kill+10
for 10
	dat     0,0
rof

stun    spl     #kill-gate,kill-gate+10
	mov     *gate,>gate
	djn.f   -1,{gate-kill-15
kill    dat     kill-gate,kill-gate+10

This one is a less used clear, MyVamp 3.7 used it. It's main advantage is
that it allows the use on djn.f, better against replicators.
I named it stargate because of the mov *gate,>gate line.

Rank    Name                    Author                   %W  %L  %T   Score
___________________________________________________________________________
  1     stargate                Anonymous                37   1  63   691   
  2     Spiral #0,2667          Anonymous                 1   4  95   98    
  3     Spiral #-1,2667         Anonymous                 1   4  95   98    
  4     Spiral #2667,*0         Anonymous                 1  40  59   62    
  5     Spiral #1,2667          Anonymous                 0  99   1   1     

As you an see it's near unable to kill spirals with an a field of zero or
small negative.

The last clear we discuss is Bjorn Guenzel's d-clear

;redcode-94
;name d-clear
;author bjoern guenzel
;strategy clear (no scanning)
;assert CORESIZE==8000
;release 23.5.96
;kill d-clear

shift           equ -3000

ptr             dat >0,4000
		dat 0,0
start           spl #0,>ptr
		mov.i bomb,>ptr
		djn.f -1,>ptr
bomb            dat >2667,bomb-ptr+3
bsource         dat 0,0
boot z for 6
		mov.i {bsource,<bootptr
rof
bootptr         spl start+shift,bsource+shift
		mov.i #0,-1
		dat 0,0

z for MAXLENGTH-CURLINE
		spl #z*100,z*101
rof


		end boot

Rank    Name                    Author                   %W  %L  %T   Score
___________________________________________________________________________
  1     d-clear                 bjoern guenzel           92   1   8   1129  
  2     Spiral #0,2667          Anonymous                 1  90   9   12    
  3     Spiral #1,2667          Anonymous                 1  90   9   12    
  4     Spiral #-1,2667         Anonymous                 1  90   9   12    
  5     Spiral #2667,*0         Anonymous                 0  96   4   4     

As you can see it's equally good against all imp species. Results may be a
little distorted because I didn't remove the boot, but not too much.

Before everyone chooses d-clear as the best one, give a look at those
results; unless you have some stunning power in your warrior you'll fall
prey of every replicator. 

Rank    Name                    Author                   %W  %L  %T   Score
___________________________________________________________________________
  1     TimeScape (1.0)         J. Pohjalainen           44  21  35   833   
  2     gate c-c2               Anonymous                38  15  47   161   
  3     stargate                Anonymous                35  35  30   135   
  4     gate c-c1               Anonymous                23  40  37   106   
  5     gate c-ca               Anonymous                11  38  51   84    
  6     d-clear                 bjoern guenzel            0  92   8   8     


______________________________________________________________________________
Questions?  Concerns?  Comments?  Complaints?  Mail them to people who care.
authors: Beppe Bezzi <bezzi@nemo.it> or Myer Bremer <bremermr@ecn.purdue.edu>
or Anton Marsden <amarsden@comp.vuw.ac.nz>
2002-2005 corewar.info. Logo C. Schmidt