Join 3,442 readers in helping fund MetaFilter (Hide)


Intel Likes the Napster Way (Wired Article).
October 31, 2001 10:56 AM   Subscribe

Intel Likes the Napster Way (Wired Article). So, we've had Napster, and its counterparts, and we've had all sorts of cheerleading for P2P. P2P has taken off in a big way in the way of IMing, and in a smaller way via projects like SETI@home. Now there's a major corp using it for internal practices in a big way. Are any of you seeing any interesting uses of P2P where you work?
posted by badstone (4 comments total)

 
At last, those Powerpoint Rangers won't clog my e-mail servers. Whew!

SETI@Home is not P2P, but distributed computing. There's a huge difference. Nobody on SETI@Home downloads stuff from other people's computers; they process data and communicate directly with central servers. Something entirely different, though both are well-served by the so-called "network effect".
posted by dhartung at 1:01 PM on October 31, 2001


Some might say Napster isn't P2P either, as it hands off some of the functions to centralized servers (as opposed to gnutella).

P2P has become one of those buzzwords that get misused so much they cease to have any meaning at all.
posted by tommasz at 1:55 PM on October 31, 2001


I attended an IEE lecture only this evening on the subject of P2P and more specifically Groove, a framework for building P2P applications.

SETI *was* included in a matrix illustrating the different types of P2P. The reasoning behind it's inclusion was that it followed the paradigm of pushing storage and processing out to the edges of the network though I'm still not convinced if it counts.

(Here goes nothing...)

DataProcessing
Index ServerNapster, AudiogalaxySETI@home, D.net
Without Index ServerGnutella

In these examples the index server is some lightweight centralised server which contains details of other clients. This seems to be the only way to combat the main problem with P2P. That is the problem of discovery, how do users initially find other users? Even Gnutella clients ended up implementing some form of index - Limewire router and the Clip2 Reflector (which recently folded) being just two examples.
posted by gi_wrighty at 3:43 PM on October 31, 2001


Looked into Groove for a while, and got very excited about it, but after I discovered that the encrypted database on my local HD was over 400 Mb with about half a dozen documents shared in two or three 'spaces' I decided to leave it alone. Horrific bloat.

Fantastic tech though, by and large, and really interesting work with security on the disk and on the wire, in particular.

Microsoft recently dumped a bundle of cash on them, FWIW.

Ray Ozzie, of Lotus Notes fame (or infamy - I am a Notes hater/Domino agnostic) is behind Groove. Read some of the Groove whitepapers if you're at all interested in corporate P2P. Exciting stuff, I think. The plan is to have it OS-independant soon too (if it's not already - I haven't played with it in a few months)...
posted by stavrosthewonderchicken at 2:56 AM on November 1, 2001


« Older Candy Bar Math...  |  "Preserving the environment is... Newer »


This thread has been archived and is closed to new comments