RE: [decentralization] FW: The first meeting of Intel's Peer-to-Peer Working Group
- On Tue, 12 Sep 2000, Lucas Gonze wrote:
> The problem is network externalities. The amount of resources available to allBut these different P2P systems don't do the same thing, and we don't even
> the subnets combined is a hell of a lot bigger than the amount of resources
> available to any one subnet. Freenet people shouldn't have to reinvent P2P
> transactions, authentication, etc if Jungle Monkey people already have these
> resources available. Ok, I can build a Radio Userland compatibility layer into
> my P2P app X, but do I also have to custom build a Gnutella layer, and a Napster
> layer, ad infinitum?
know if they work!
Once there are multiple P2P systems which (A) do similar things (B) which
people actually use and (C) scale, then let's talk about standards. I
don't think we're at that point yet.
(If some P2P developers want to borrow ideas/protocols/code from each
other, I'm all for that. But let's not confuse that with standardization.)
Wesley Felter - wesf@... - http://www.cs.utexas.edu/users/wesf/
> But these different P2P systems don't do the same thing,taint so. They are extremely similar. Routing algorithms differ a little, but
not that much. Napster, for example, is just a simple 3-node Gnutella network.
And the general problems are the same for every P2P network - reputation
management, throttling, decentralized transactions.
> and we don't evenOr maybe we can be sure that they don't: http://dss.clip2.com/dss_barrier.html
> know if they work!
from that analysis:
Quantitatively, we have observed a correlation between the average query rate
and the responsiveness of the network to Gnutella pings. As the average query
rate increases, the fraction of hosts returning Gnutella pongs in response to
Gnutella pings decreases.
I suggest that this is because incoming messages cause the message queue to spin
faster, which decreases the amount of time available for responses to return and
be passed back. This problem occurs with all queue-based routing. Again, it's
a common problem.
- At 11:00 AM 9/12/00 -0700, Lucas Gonze wrote:
>Info, slightly trimmed, on Intel's working group. Note that this is a pay toWell, hey. News.com is doubting whether there's money in P2P anyway, so
>play event, meaning non-commercial developers are out of this dialog for the
who'd want to pay to sit in a room anyway?
Oh well. Money, rooms, whatever.
XML Elements of Style / XML: A Primer, 2nd Ed.
XHTML: Migrating Toward XML
http://www.simonstl.com - XML essays and books
- Wow, the hype is dying before people could even release any
products *chuckles* too funny. I wonder if the press is
getting just too efficent, and if there is room to hype
something AND make a product actually work before the
next-big-thing drains everyone away.
And the Intel meeting got MOVED?
Think they will start kicking people out?
--- In firstname.lastname@example.org, "Simon St.Laurent"
> Well, hey. News.com is doubting whether there's money in P2Panyway, so
> who'd want to pay to sit in a room anyway?http://news.cnet.com/news/0-1005-200-2758911.html?tag=st.ne.1002.bgif.
- Hey Adam, nice to see you here! Maybe they're thinking of moving the
meeting to Minneapolis to accomodate us. :P If not we should just go grab
a beer because I don't think I've seen you since rootfest '99...
On Tue, 19 Sep 2000 beberg@... wrote:
> Wow, the hype is dying before people could even release any
> products *chuckles* too funny. I wonder if the press is
> getting just too efficent, and if there is room to hype
> something AND make a product actually work before the
> next-big-thing drains everyone away.
> And the Intel meeting got MOVED?
> Think they will start kicking people out?
> --- In email@example.com, "Simon St.Laurent"
> <simonstl@s...> wrote:
> > Well, hey. News.com is doubting whether there's money in P2P
> anyway, so
> > who'd want to pay to sit in a room anyway?
> To unsubscribe from this group, send an email to:
- Yea, only I moved too...
The meeting _was_ not even a mile from where I live now.
Obviously it's grown to more poeple then can possibly agree
on anything, I doubt the Intel group will even agree
that TCP/IP is the way to go.
Should be a riot.
--- In firstname.lastname@example.org, Justin Chapweske <orasis@a...>
> Hey Adam, nice to see you here! Maybe they're thinking of movingthe
> meeting to Minneapolis to accomodate us. :P If not we should justgo grab
> a beer because I don't think I've seen you since rootfest '99...