Yesterday, Tim Bray posted Bye-bye Home Page? with some perceptions of how RSS feeds are overtaking use of the browser's default home page, and also overtaking the use of website home pages. (I came across this via: Steve Gillmor mentioned Tim's post in his blog, which was mentioned in Dan Gillmor's blog. All recommended.)

Tim says in his post:

So, to the extent that you have a dynamic presence and an RSS feed, your homepage doesn't matter that much. Which means, I guess, that while ongoing's front page gets more traffic than any non-slashdotted individual piece, that probably won't last. Which means that if you want people who land in your Web territory to look sideways at anything other than where they landed, dressing up your home page just isn't going to help.

I know many of the large websites I have worked on ended up having 70%-80% or more of their entry traffic coming in from pages other than the homepage, i.e., through web searches and inbound links to pages deep within the site.

With blogs, because of blogrolls I am assuming, there is a lot of linking to blog home pages that drives traffic to home page. But, other than this type of link, I would guess that deep linking into blogs would show a similar pattern as with large websites: most links and traffic go deeper into the site than the home page.

With RSS however, the "bag" that holds together a website, is pierced. That bag is the combination of URLs and a visual design that makes a bunch of otherwise completely independent web pages *appear* to be a website. (See also, There's No Such Thing as a Web Site, also by Tim Bray.)

With RSS feeds, people get website content outside of that bag. It is not just that they get it on their desktops, but sites / services that read RSS feeds are caching and/or aggregating content in ways where the so-called website (where the content originates) takes a back seat to the content itself.

Website retain at least two purposes that RSS does not address: 1) a URL "starting point" for a particular frame of reference, and 2) an information architecture and (user) experience for a particular frame of reference. So, for example, if you want to learn about the iCite net, this website exists to fulfill those two purposes, in particular, expressing my (the iCite net founder's) frame of reference about the iCite net.

I think it is quite possible for these purposes of websites to be addressed in the evolution of RSS, dependent really upon people making every aspect of their websites accessible through RSS feeds.

the iCite net may function as a kind of bridge between these two purposes and the purposes of RSS with regards to aggregation and "feeding" on information. iCites are a way to make every aspect of a website accessible through RSS feeds, including the site's relationships with other site's.

iCites provide URL starting points, but they are starting points to collections (architectures) of information, whether they are websites or beyond websites (either in the sense of being metadata about websites or in the sense of connecting together more than one website).

An iCite can be a way to define a website, an iCite can be a way to join information without regard for the website bag, and an iCite can be a way to join a website with information outside of it.

iCites also provide a means to package information architecture and aggregate it. For example, I came to Tim's post today by going through: my desktop RSS reader to a post on Dan Gillmor's blog to a post on Steve Gillmor's blog to Tim's post on his blog. Tim doesn't include his full post in his RSS feed, so I am forced to read it on his site. But, that is good in that Tim's site has features that aren't in the RSS, like useful ways to navigate to related posts.

With iCites, it will be possible to aggregate navigation to related posts along with the posts. The navigation information aggregated can span multiple sites (like viewing Dan's, Steve's, and Tim's in a single feed), and/or be representative of a single site (like viewing Tim's related post navigation). As with RSS, people will be able to subscribe to the parts they want.

I don't think websites will go bye bye--homepages won't go bye bye either. But, I think the focus will continue to shift towards de-composing information and making it package-able and accessible via many decentralized interfaces, vs the websites as a singular centralized bag. I also think that many of the interesting things about information architecture, user experience, and branding will find expression in these small packages and in their loose joining.

I imagine that, relatively coincident with this, people will continue to find ways to make sites more like environments, more like places--large sites more like cities. Every large site may have a "library" or a "radio station" with similar information organization features based on providing access to information packages, but the aesthetic quality of those environments will evolve differently from each other, as a reflection of the way people (users) individually and collectively want to feel or express themselves.