On Apr 23, 2008, at 5:59 PM, Henry Sinnreich wrote: > Another option is for the terminology paper by Dean Willis et al. to > be > updated with an updated and comprehensive glossary for P2P.
Sure -- bring vocabulary words to the list, and once we have agreement on what they mean, we'll stick 'em into the draft. That might not be as easy as we think, and it's part of why everything currently has a "p2psip" label in front of it. A P2PSIP overlay is not a BitTorrent overlay. For example, there's no concept of discretely bounded namespaces in BitTorrent. And as far as I know, BitTorrent has no concept of "resource discovery" -- we use web sites to distribute the resource keys. But we still think of BitTorrent as P2P and an overlay . . . So some of our terms are going to have names that are used in other literature to mean something that's not quite exactly the same thing. As long as we have them prefixed with "p2psip", it's clear that we're referring to "our" definition. Presumably, we understand this and can drop the "p2psip" part from the names, but still know it is there. But eventually some joker is going to pop up and say "You can't call this thing a Node, because a Node is another name for a Peer, and it's not a Peer, it's a Client" or some such, and I'm going to have to hit him with a blunt object. That's more-or-less already happened, and the blunt object we used was the concepts draft, with its mind-numbing repetition of the "p2psip" modifier. -- Dean _______________________________________________ P2PSIP mailing list [email protected] https://www.ietf.org/mailman/listinfo/p2psip
