the stuff is kind of hard to get into the head. more than 60 working groups (wgs) are talking about specific problems: routing, gateways, protocols, standards that are in the draft and so on. ipv6, dns, security, multimedia, enum ...

the dream to feed a toaster with bits and bytes is not easy. there is no way anymore to trig a lever and switch to a new protocol or transporting system as it was still possible 20 years ago with introducing tcp and dns.

listened to a talk about the outline of a new one: dccp, datagram congestion control protocol - something that lies between udp and tcp, they say. somehow they were talking about the goal to have a protocol that allows unreliable flow of junkets of data so that it would be more suitable for applications such as streaming media. if it ever comes true it will not be before next year that the final call will be out.

the only interesting thing for me was to watch the process. wouldn't get more anyhow. there was a lot of criticism and suggestion to look more deeper into that and this, but i guess the most important statement was: the whole stuff would change more of tcp than of udp. but to do so, it must be a lot simpler and with less features - even they look and sound nice, the big heads add.