¶ distributed systems, stata move Posted 21 years, 10 months ago on March 18, 2004
We're moving to our brand new building this weekend. Consequently, the machines hosting this blog will be down for a couple days.

The more I learn about it, the more I feel that distributed and networked systems is a field that's been studied to death. Well, not to death, but virtually all the low hanging fruit has been taken. It's not to say there's nothing left in the field, just that relative to the number of people in it (a looooooot) there's not much in terms of deep research problems.

For my class, we have to do a final project. phds are expected to do an original research project. So my group was just sitting around, trying to think of stuff to do and it was just painful. We were like "distributed file systems... nope... shared memory... nope... distributed spam filters... nope... peer to peer systems.. ew..."

So there's a lot of stuff that's wrong with the internet today, but it's not because solutions haven't been found. It's that so much is invested in the existing infrastructure that nobody wants to switch. IPv6 is the classic example. Not enough IP addresses to go around, but nobody wants to switch to a new system so we get these kludges like NAT and CIDR. SMTP is another. It's ridiculously easy to spoof someone's mail address, and spam people.

Maybe I'm just being pessimistic. But then I think of fields like vision and speech, and it's clear that there are so many deep and open questions there, like generalized object recognition, and scene reconstruction.

I think we're gonna end up doing some peer to peer thing anyways, cause we can't think of a better idea. something about enhancing bittorrent to replicate trackers and integrate it into searchable, stable networks like edonkey2k.

---

No comments, be the first!

---

Comments disabled until the spammers go away. I hope you comment spammers all die horrible deaths and are forced to delete endless streams of comment spam in your days in purgatory.