But what I respect about the IETF is that it only adopted enough process to keep things moving -- focusing at all times as much as possible on substance over process, including "fairness" and "democracy." In the later chapters of the book, I argue that fixing the Net is ours to do, in just the informal, junior, unchartered way the IETF operates (those adjectives taken from the RFC entitled "30 Years of RFCs").

Your warning does not hold water! If what you propose comes about then why didn't the same thing happen to television when the Internet was born?

Not sure I understand your emphatic and slightly peeved claim. You mean, why didn't the Internet kill television? Because it's only recently that moving images of TV quality were possible on the Net, mostly with the publishers who fill TV's content objecting every step of the way to making it happen. Ultimately I do worry that TV *will* kill the Net -- that we'll end up with something approximating the ethos and structure of Cable TV, just done over TCP/IP.

Do you make any distinctions between Wi-Fi and cellular transports?

Well, I respect the medium-independence of TCP/IP hourglass architecture. I'm with fellow travelers like Tim Wu on wireless Carterphone and am heartened by Android and by Verizon's statements about opening up its network. The devil will be in the details, and of course they won't (rightly) want to open up their networks so much that they lose the advantages of managed networks, such as user authentication that allows for billing, abuse complaints, etc.

AOL began life as a separate system that was later linked to the Net. Even before it was linked, (even in the days when it only ran Q-Link for Commodore 8 bit machines) there were predators. The bigger it got, the more predators it drew. Won't any successful walled gardens of tomorrow face the same problem?

Yes, I think they will. I just don't want them to be predators of the same ilk. Remember when it was AOL v. CompuServe v. The Source? (BTW, I recommend heartily A Game of Thrones.)

It's a Catch-22 - we need standards to make our lives easy, but speed of standards release can't catch up with the speed of changing technology.

Yes - if we can preserve the experimentalist architecture for everyone, while allowing people to roll up their sleeves and get their work done when they need to, it'll be easier to let the forking and restandardizing to take place.

The majority of PC users aren't hobbyists. They don't tinker with their systems, create applets or participate in wikis. They just use their PCs to access Internet content created by someone else. Isn't it OK if these people migrate to appliances?

Use appliances, sure; migrate, no. Skype is that much more useful as more people use it -- not just nerds. It's good to be able to Skype grandma. The generative PC is (well, was) useful to non-techies because they could install software written by others. This serves as an important safety valve -- if TiVo gets too grabby or over-regulated, we can revert to MythTV, so long as the PC infrastructure is there to support it.

So your world is defined by expediting "standards" and a lot more cross licensing?

I worry that as soon as we're into the zone of cross-licensing -- that's already an L-word that stops molecular motion -- it assumes there's IP [intellectual property] to be licensed to begin with. I'd like to make sure we maintain a hardware infrastructure that allows nerds to come up with new stuff on their own and deploy it to the rest of us -- a safety valve against the more formalized/proprietary systems that will naturally be competing too.