Archived from groups: rec.games.corewar (More info?)
All,
It's unfortunate that I won't be able to attend the 20th anniversary
meeting in Berlin. I would have loved to meet in person all the other
Corewar fanatics in our small but lively community. But in lieu of
that, I'd like to share this bit of Corewar history with everyone, in
honor of the occasion.
I must have been one of the first people who started playing Corewar,
in 1985, and (as some of you may know) was the winner of the very
first Corewar tournament, in 1986. It's hard to imagine now, when we
are all so connected by email and the Internet, how important that
first tournament was. Up until that time, all of us pioneers had been
coding in almost complete isolation, with just a few bits of
snail-mail correspondence starting to flow as the ICWS formed. Other
than Imp and Dwarf, I didn't have access to a single warrior written
by another person! Corewar was almost like solitaire, rather than a
competitive game.
So it was a pleasant surprise to see how widespread Corewar had
become, when the first competition received 31 entries from 17 authors
in 4 countries. Here is a portion of the letter I got announcing the
results:
"Well, the First International Core War Tournament is history.
Thirty-one (31) entries arrived of which twenty-seven (27) survived
assembly to do battle. Battle programs employing every conceivable
strategy (well, almost) 'duked it out' inside of a row of AT&T 6300's
of the Computer Museum in Boston for two days. When the screens
cleared, a program named 'MICE' stood at the top of the scoreboard.
Across the globe, anxious authors awaited the results, already
plotting next year's strategies - tomorrow's warriors.
"The competition took place on Sept 12th and 13th [1986] between
entries from the United States, Canada, West Germany and Japan, making
it truly international.
"In personal attendance were Core War originator A.K. Dewdney, ICWS
Director Mark Clarkson and his wife and acting ICWS secretary Beth
Clarkson. Saturday, the day of the finals, saw the Computer Museum's
auditorium playing host to 75 or so spectators. Crowds of up to 25 at
a time sat and watched the progress of the programs marked by colored
dots marching across a row of monitors. Up front, Dr. Dewdney provided
color commentary.
"Due to the number of entries received and the time constraints, first
round competitions were run in 2000 word cores for a maximum of 20,000
instruction cycles. Some programs were designed to run in larger core
sizes and did not fare well in the smaller space. For some programs,
self-destruction was the order of the day.
"The programs were divided arbitrarily into two divisions - I and II.
All programs were run twice against every other program in its
division. Losses were subtracted from wins to score - ties were not
counted.
"The top four (4) programs in each division were then moved to the
semi-finals. Every program was run against every other four times in
4000 word cores.
"When it was all over, the finals looked like this:
"1st place - Mice by Chip Wendell
"2nd place - Chang1 by Morrison Chang
"3rd place - Midget by Chip Wendell"
The letter went on to list the scores from each round, and the source
code of the eight semifinalists. Mice was the only warrior that used
infinite replication as a strategy, so apparently it was the world's
first paper.
It's truly staggering to see how far Corewar has progressed in the
last 20 years. Redcode has expanded and matured into a very powerful
instruction set. The warriors of today are complex and vicious
creations against which Mice, the best of its day, would not stand the
slightest chance of survival. MARS simulators are freely available
that could run all 516 battles of the entire 1986 tournament in two
minutes, instead of the two days it took back then. (If you want to
get an idea of how it looked to the audience of the '86 tournament,
set the speed of CoreWin - I don't think pMars goes that slow - to 100
instructions per second, the third-slowest setting.) Even the name has
evolved - what was originally Core War became Core Wars (although that
was never universally accepted), and now is just a single word,
Corewar.
But I think that the most significant change over the last 20 years is
in how tightly connected the Corewar community has become. Ideas and
warriors flow like water over the Internet, and we are far more
acquainted with each other than the members of the ICWS ever were. I
may not be attending the anniversary meeting, but I'm sure that, after
I get through reading all the posts and logs that are sure to come,
I'll feel like I was almost there in person.
Any other old timers care to share some history/reminiscences?
- Chip
All,
It's unfortunate that I won't be able to attend the 20th anniversary
meeting in Berlin. I would have loved to meet in person all the other
Corewar fanatics in our small but lively community. But in lieu of
that, I'd like to share this bit of Corewar history with everyone, in
honor of the occasion.
I must have been one of the first people who started playing Corewar,
in 1985, and (as some of you may know) was the winner of the very
first Corewar tournament, in 1986. It's hard to imagine now, when we
are all so connected by email and the Internet, how important that
first tournament was. Up until that time, all of us pioneers had been
coding in almost complete isolation, with just a few bits of
snail-mail correspondence starting to flow as the ICWS formed. Other
than Imp and Dwarf, I didn't have access to a single warrior written
by another person! Corewar was almost like solitaire, rather than a
competitive game.
So it was a pleasant surprise to see how widespread Corewar had
become, when the first competition received 31 entries from 17 authors
in 4 countries. Here is a portion of the letter I got announcing the
results:
"Well, the First International Core War Tournament is history.
Thirty-one (31) entries arrived of which twenty-seven (27) survived
assembly to do battle. Battle programs employing every conceivable
strategy (well, almost) 'duked it out' inside of a row of AT&T 6300's
of the Computer Museum in Boston for two days. When the screens
cleared, a program named 'MICE' stood at the top of the scoreboard.
Across the globe, anxious authors awaited the results, already
plotting next year's strategies - tomorrow's warriors.
"The competition took place on Sept 12th and 13th [1986] between
entries from the United States, Canada, West Germany and Japan, making
it truly international.
"In personal attendance were Core War originator A.K. Dewdney, ICWS
Director Mark Clarkson and his wife and acting ICWS secretary Beth
Clarkson. Saturday, the day of the finals, saw the Computer Museum's
auditorium playing host to 75 or so spectators. Crowds of up to 25 at
a time sat and watched the progress of the programs marked by colored
dots marching across a row of monitors. Up front, Dr. Dewdney provided
color commentary.
"Due to the number of entries received and the time constraints, first
round competitions were run in 2000 word cores for a maximum of 20,000
instruction cycles. Some programs were designed to run in larger core
sizes and did not fare well in the smaller space. For some programs,
self-destruction was the order of the day.
"The programs were divided arbitrarily into two divisions - I and II.
All programs were run twice against every other program in its
division. Losses were subtracted from wins to score - ties were not
counted.
"The top four (4) programs in each division were then moved to the
semi-finals. Every program was run against every other four times in
4000 word cores.
"When it was all over, the finals looked like this:
"1st place - Mice by Chip Wendell
"2nd place - Chang1 by Morrison Chang
"3rd place - Midget by Chip Wendell"
The letter went on to list the scores from each round, and the source
code of the eight semifinalists. Mice was the only warrior that used
infinite replication as a strategy, so apparently it was the world's
first paper.
It's truly staggering to see how far Corewar has progressed in the
last 20 years. Redcode has expanded and matured into a very powerful
instruction set. The warriors of today are complex and vicious
creations against which Mice, the best of its day, would not stand the
slightest chance of survival. MARS simulators are freely available
that could run all 516 battles of the entire 1986 tournament in two
minutes, instead of the two days it took back then. (If you want to
get an idea of how it looked to the audience of the '86 tournament,
set the speed of CoreWin - I don't think pMars goes that slow - to 100
instructions per second, the third-slowest setting.) Even the name has
evolved - what was originally Core War became Core Wars (although that
was never universally accepted), and now is just a single word,
Corewar.
But I think that the most significant change over the last 20 years is
in how tightly connected the Corewar community has become. Ideas and
warriors flow like water over the Internet, and we are far more
acquainted with each other than the members of the ICWS ever were. I
may not be attending the anniversary meeting, but I'm sure that, after
I get through reading all the posts and logs that are sure to come,
I'll feel like I was almost there in person.
Any other old timers care to share some history/reminiscences?
- Chip
Facebook
Twitter
Instagram