The death of NNTP
You can take a user to a ‘net protocol, but you can’t make them use it.
Protocols are being blurred like crazy at the moment, and it’s not clear what’s going to be left when the dust settles and the metaphors are unmixed. Google’s new gmail service is basically web-based (i.e. HTML across HTTP) e-mail (typically SMTP), but with threads concatenated in a way that makes them look like Usenet (i.e. NNTP) discussion threads, or blog (i.e. RSS/Atom/whatever-ends-up-as-convention) entries. Google’s own Usenet archive (i.e. NNTP feeds presented as HTML across HTTP) is now being experimentally presented as an Atom feed (i.e. like a blog).
Something’s going to be left behind, and I have a sneaking suspicion that it’s going to be NNTP. Since the web took off, it’s been an increasing puzzle to new ‘net users just what Usenet is. People call discussion groups ‘sites’, and ask what their ‘address’ is – by which, of course, they mean the URL – or the ‘URI’ these days, the convention-change also a signal of the blurring between protocols. (And it doesn’t really serve to tell them that the address is news:name-of-group.) What people know, and what they’re comfortable with, is the web. The software they’re most comfortable with, is a web browser.
I think we’ll be left building most things on top of HTTP. It’s really happening already. I’ve no idea what ‘news’ syndication standard (RSS, Atom) will prevail. It doesn’t really matter for this discussion, since they’re both mechanisms which use XML (which is the direction HTML is heading anyhow) across HTTP. Again, it doesn’t matter which RSS/Atom aggregator software prevails. It won’t be long before the functionality of an aggregator becomes integrated with existing browers. In the end, a blog, or a news feed which uses RSS/Atom as its form, is just a web-site which updates regularly, and in a particular direction. That’s how they’ll be seen, and it’s how it makes sense to see them.
Reading bsag’s discussion of the relative merits of various Mac OS X news aggregators, I was struck by her calling them ‘news readers’. It’s a good description, but it’s a startling usurping of the name that has typically been used for Usenet (i.e. NNTP) readers. The popularity, and newness, of blogs and blog-like feeds, seems unstoppable. When something loses its name, its days are probably numbered.
Not that this means that Usenet’s days are numbered. Despite the diversification and schisming of ‘net discussion across web-based message boards, blogs and such, Usenet is still a healthy medium. It’s NNTP that’s doomed. The popularity of Google as a Usenet interface is chipping away at both NNTP’s use, and its visibility. Usenet is increasingly seen as something which happens on the web.
It won’t be so long before news aggregators become part of web-browsers, and it’s only a small step from that to the swallowing up of Usenet. In the end, there isn’t anything fundamentally unique about Usenet’s structure which merits its own protocol any more. Usenet groups look an awful lot like communal blogs, and a parsimony of protocol use, along with a user community which wants to bring separate streams together into a similar (if not the same) interface, will cause any distinction to narrow, and then vanish. There’ll be individual news feeds (blogs), communal but private news feeds (group blogs), and communal, public news feeds (Usenet). But the mechnisms will end up being much the same.
There are certainly technical issues that I’m not even beginning to consider – such as the natural site-mirroring process that’s a fundamental part of Usenet, and which distributes its load and access – but those will be overcome. Something that’s good, that’s valued, finds a way to persist, and that’ll be true of Usenet, whether it uses NNTP, or RSS/Atom on top of HTTP, or some protocol yet to be invented. Usenet’s value isn’t in its protocols. Its in the content, and community, and there’s quite enough of that for it to continue to thrive.