No one is getting Web aggregation quite right. That's one of the big take-aways from "Web Aggregation: What Works, What Doesn't," one of the breakout sessions at the ReadWrite Real-Time Web Summit.
We first heard about the fire hose meme several years ago in discussions about RSS. It was often used as a way to describe how information comes to you in a feed. The context has changed as real-time data becomes pervasive, and the questions about its volume persists.
The fire hose conversation is often centered on Twitter these days, but it's an issue across the social Web. Perhaps most of all, we should be thinking about what are the subsets of the fire hose and, in particular, how we use data streams in our lives.
The real-time Web ebbs and flows. Most people find the real-time information well after it has been published in an activity stream. Thus, a proliferation of new search engines are coming to market, looking to capture this real-time data and making it relevant to users.
Another distinction made in the discussion centered on how we consume real-time information and the persistence with which we need to get it.
For instance, some information you do not need to be pinged on every 30 seconds. Instead, it may make more sense for it to be pushed to you when you need it. You may only get this information on rare occasions, such as an emergency. So when you do get that information, it is very relevant.
The real-time Web may be instant, but our lives do not work that way. One participant said that he may be interested in a photo of his son that appears in his stream but perhaps not the photo of his son's buddy, who happens to like Scandinavian death metal music.
So, the question becomes, how will the real-time Web fully develop. For Joseph Smarr of Plaxo, that's where open standards come into play. Interestingly, open standards are emerging as an oft-discussed issue at the Summit.
Smarr made the point that RSS and Atom were designed to share the titles and bodies of blog posts. What we are actually sharing in an activity stream is far richer. What we need is language that embodies the far richer meta data that comes in a real-time activity stream. Pubsubhubbub and RSSCloud are starts, but there is still a lot of work to do to put the pieces together.