Been thinking alot about what’s needed to pull many of the emerging tools and applications together. In that spirit, Phil Wolff raises some good questions once again, sent via Marc Canter’s blog:
“How do you scale SyndicationSpace to hundreds of millions of feeds being checked all the time? The answer is intermediaries. Proxy caching, P2P, central aggregators, etc. Don’t know which blend of technologies is the right one, but they all require some form of
(a) authentication, so you can trust that this copy of the feed come from the source you know,
(b) encryption (optional), so nobody along the way can read/modify the content of a private personalized feed, and
(c) checksum or another method of assuring that what was received is identical to what was sent.
This tastes like public key infrastructure.
Another application: Blogspace isn’t flat. FOAF data, whether explicit via blogrolls or implicit in hyperlinks, should be very useful in P2P. Your subscription list probably has many feeds in common with the people in your weblogNeighborhood, Technorati cosmos, or other people in your blogroll. While John Udell or Marc Canter may be physically distant (in traceroute and ping times), the chances may be as high as 25% that they have cached at least one of my feeds. As long as the feeds have clear BestIfUsedBy/ Expiration dates/ times, the social network can make optimization very efficient.”
Been thinking alot about what’s needed to pull many of the emerging tools and applications together. In that spirit, Phil Wolff raises some good questions once again, sent via Marc Canter’s blog:
“How do you scale SyndicationSpace to hundreds of millions of feeds being checked all the time? The answer is intermediaries. Proxy caching, P2P, central aggregators, etc. Don’t know which blend of technologies is the right one, but they all require some form of
(a) authentication, so you can trust that this copy of the feed come from the source you know,
(b) encryption (optional), so nobody along the way can read/modify the content of a private personalized feed, and
(c) checksum or another method of assuring that what was received is identical to what was sent.
This tastes like public key infrastructure.
Another application: Blogspace isn’t flat. FOAF data, whether explicit via blogrolls or implicit in hyperlinks, should be very useful in P2P. Your subscription list probably has many feeds in common with the people in your weblogNeighborhood, Technorati cosmos, or other people in your blogroll. While John Udell or Marc Canter may be physically distant (in traceroute and ping times), the chances may be as high as 25% that they have cached at least one of my feeds. As long as the feeds have clear BestIfUsedBy/ Expiration dates/ times, the social network can make optimization very efficient.”