Last week, in DOAP and the Next Advogato, I touched on my desire for sites that allow me to use my own website as the authoritative source of information while at the same time publishing that information and aggregating it with that from others for the purpose of searching, rating, etc.
In Some 'Web as platform' noodling, Kottke ponders the opposite: putting different components of your digital life out on different websites: using one site for your photos, another for your playlists, another for your calendar, etc.
But this is exactly what I don't want to do. I'm happy to publish my blog and open source project info on Advogato but I'd like it to act purely as an aggregator of my RSS/Atom and DOAP feed, not the authoritative source. I'm happy to use LinkedIn but I'd like LinkedIn to poll my FOAF - not have LinkedIn be the authoritative source. I'd like to be able to publish events from my calendar to Upcoming.org rather than enter them in Upcoming.org and use their feed if I want to use my own info on my site.
I think there's a lot of value in pure aggregation sites. I like the value-add that reading blogs in Bloglines provides in terms of references and recommendations. But I'm not interested in pulling my blogroll from them, I'd rather push it to them (or have them pull it from me).
The value of a LinkedIn or Upcoming.org is in the aggregation, not using them as an authoring tool or repository for one's own data. They should focus on competing on the value-add of their aggregation. I don't see any disadvantage for them in opening up the input mechanism to pull the source information from external authoritative feeds (or support the information being pushed to them).
I'm not ruling out the need for information hosting services. But I think aggregators and hosting services are different beasts and separating them provides many advantages to both providers and consumers of information.
UPDATE (2004-09-25): Now see More on Aggregation Versus Hosting