Date:	1997-02-20
		From:	John
		Subject:	RE: Alexander

It's been a frantic day here, so I've only had a few seconds to read over your note. My initial reaction is that I am skeptical about all of the variations you propose because I *think* the behavior of choosing from the spark list at random is the very essence of the algorithm - muck with that and all bets are off. The stack proposal, I think, would result in the first spark growing like a cancer; the second spark would not be touched until the first was fully contained. The queue proposal is better but might result in more uniform growth since each spark would expand outward in a uniform way like ripples in a pond and the interference between ripples would proceed in a relatively uniform manner. I can't predict with confidence what would happen under both models, and agree it would be interesting to try them and see.

Once upon a time I did have a Hypercard stack which would plot a file of 1s and 0s (or was it 1s and 2s?) - chances are good that I can find it. I can then take a screenshot, pop it into Photoshop, and let the colors fly! I will be glad to email you a BMP (or whatever you require).

I will look forward to seeing your maps!

Yours in haste,

Epicurious