Anil Dashes gets it absolutely correct when he says of Web 2.0: "But this stuff's too important for us to let it become another bubble with no editing and no filtering." Indeed. The previous bubble put a lot of people off Internet investment for a long time. The Web 2.0 field is good in as much as they are great applications. All the bullshitting the VC's, talking bollocks, slapping "beta" on everything and having this damn invite system. 
All the Web 2.0 "theorists" (from Tim O'Reilly on down) describe how it's an interactive and emergent system that values the user in his wholeness. To which I say, Bollocks.
How is the user valued if, when he wants to try your website, he has to go searching on Technorati to find an invite code, or visit blogs which he doesn't care about in order to get an invite?
This is exactly the thing with CoComment. I found out about CoComment at six in the evening on Monday night. I spent maybe twenty minutes pissing around trying invite codes that were posted on Scoble's blog and the various linked sites, and then poking around trying to find them using Google BlogSearch.
This is to let me have a look at CoComment to see if it's any good or not. Eventually I register for an invite code and sign up at midnight. By which time I'm going to bed, and so don't discuss it on my site. What purpose has the invite codes served? Absolutely none.
If Web 2.0 applications gain value through users, then the Web 2.0 companies need to value their users. Forcing them to jump through hoops to play with a public beta is an insult to their time.
As is not providing decent export options. If your service is good, you've got nothing to fear from letting people leave.
The battle is half done, since the guys sitting near me on the train have just been discussing their blogs (although they mean journals).
