Re: [life] Real-time Life Proposal (v1.1)

[Nick Waterman <nick-pbmlife@nilex.co.uk>]

Eric Shultz wrote:
> I'm reproposing Real-Time life.  This proposal is much the same as I
> suggested before.  I've changed positions on the three different color
> reproduction debate, thanks to Noseynick's proof that a neutral cell
> would not take over the board.

It wasn't so much of a proof that it was IMPOSSIBLE, just that I was
fairly convinced that it wasn't very likely, using the rules that I'd
chosen to test.

> I've seen a couple of people offer to
> code the game, but haven't done a gamerz module before.

That includes me, and I've even found some useful source code on
ftp.gamerz.net, but I've not had time to look at it, let alone write
anything for it. I'm sorry, I'm concentrating too much on my
"Interstellar Intrigue", which I've chosen to do as a web CGI program
before experimenting with pbmserv. Oh, yes, I'm hoping to have the II
CGI in a "testable" condition within a week or so, but I'll let you all
know... anyway, back to the point...

> Number of players - Theoretically infinite.  Practically, I
> would say 26 (each gets a letter of the alphabet), or
> however many letters, numbers, and symbols can be displayed
> on everyone's email.

You could immediately double it by simply using upper AND lower case,
but in practice, I suspect that even 26 players at a time is likely to
be WAY too chaotic.

> Playing field - Again, theoretically infinite.  Practically,
> it should be as wide as an email message.  The length could
> vary according to number of players.

You also have to decide what to do with gliders that go off the edge, if
anything. On a board that's non-infinite, and is surrounded by squares
that are "fixed dead/blank/whatever", they'll usually hit the edge and
turn into 2x2 squares... which is fine... whatever...

> Start - players can start anytime, anywhere (which may
> require a "join" command in addition to "Challenge").  A
> player gets 10 cells to start.  They can place those 10
> cells anywhere on the grid, as long as they are touching
> (horiz, vert, or diag)

So you can't place 2 seperate groups on opposite sides of the boarad?
Again, fine, whatever...

> Each real-time hour, a game tick goes by. You get movement
> points each tick, based on the number of cells you have.

Incidentally, until I've actually looked, I've got no idea how possible
this is within the pbmserv "universe". AT THE MOMENT pbmserv only ever
"does something" in response to an email - it has no "ticks". It MIGHT
be trivial for richard to give us a cron-job, but don't assume it's
going to be that easy even after he does.

> You get one movement point for every 10 cells (never less
> than 1).  You can use the movement points to move cells.

I would have opted for creating or destroying cells of your own, myself,
but moving them is probably fine. This kind of variation can probably be
easily added into "-options" stuff later. Certainly writing A WORKING
ENGINE is the main thing to worry about first.

> If an empty space has three neighbors, it MAY reproduce.
> This is a slight change from classic rules.  Read on...
A:
> If an empty space has three neighbors all from the same
> player, it reproduces a cell for that player.
B:
> If an empty space has two neighbors from Player A, and one
> neighbor from Player B, it reproduces a cell for Player A.
> (2 cells beats just one)
C:
> If an empty space has one neighbor from Player A, one from
> Player B, and one from Player C, a neutral cell is reproduced.  It has
> been proved fairly conclusively that a neutral cell will soon die out
> and not take over.

No, sorry, you've not said enough about neutral cells, and what they DO
count towards and what they DON'T. the rules I used were:

If an empty space has 3 neighbours:

  3 of one player creates a cell for that playey
  3 of neutral creates a neutral.
  2 neutral and 1 of player A creates a cell for player A, even though
neutral outnumbers her.
  2 of player A and 1 of player B creates a cell for player A.
  2 of player A and 1 neutral still creates a cell for player A.
  3 owned by 3 different players creates a neutral square.

I think that covers all the posibilities fairly un-ambigously, when
combined with your rules for 0,1,2, 4,5,6,7,8 neighbours.

> Immediately after each generation, a player gets points
> added to their total score.  Each generation, a player gets
> the number of cells they have, divided by the number of
> turns they have been in the game. Yes, this means a player
[snip]

I'm not sure how well this scoring would work, but on the other hand the
only REAL way to find out would be to try it quite a bit and decide how
fair it is, and what might be done to improve it.

-- 
Nick Waterman. Senior Systems Administrator, Sony Interactive Services
mailto:nick-sig@nilex.co.uk   http://www.nilex.co.uk/~nick/
AX25:G7RZQ@GB7DEO.#32.GBR.EU   #include <stddisclaimer>  Team *AMIGA*!
Get a Clue.  Buy an Amiga you schmuck!