This is a response to volume 4, Stuart Card Number 4 ("re John's Explaination")...
Thank you for your kind words. I was rather pleased with my little mini-essay on programming. It is very pleasant to explain things of personal importance and very satisfying to get ideas out onto the undying page. That's what Archipelago is all about.
Which leads me to a request: could you say a few words about the technical side of poetry? Most of us have a vague idea what an iambic pentameter is, but many of the subtleties escape us. Why is this stuff important? How does the constraint of a certain pattern affect your creative process? Can you describe what it's like trying to fit words into their "halters" (a word you used in your introduction to the Bestiary)? How do you decide what scheme to use? What do YOU hear when you listen to poetry?
I thought your review of the Bat Poet was one of the best descriptions of poetry I have ever run across. Now that I've got Holly churning away on music theory, perhaps I can coax you into a little poetry theory (the two are somewhat related).
You also asked how I would go about making a computer out of waterfalls. Actually, it's always been an ambition of mine to try to design such a computer. I've tossed around the idea of a water-computer in several of my stories, and once I spent some time reading up on 18th century water driven mechanisms, but I've never really come up with a clear conception of how such a thing would work.
There's no doubt that it could work in theory. A computer is simply a symbol manipulator, that is, a device that stores and changes symbols. And ANYTHING can be used to represent a symbol, including water pressure or tank volume (in an analog machine) or cups which are either empty or full (in a digital machine).
I don't envision something with a lot of valves and pipework, though. I'd like to use free flowing cascades of water and profound "memory pools." The input and output devices baffle me, although it would be nice to use wind and Aeolian Harps to whisper the symbols.
Such a computer would be of little interest to IBM. It would be large, unwieldy, and VERY slow. All in all, the perfect Ponarv!