orkut article on c|net

Mostly content-free article about Orkut on News.com. It does mention that there are only three engineers working on the system, but most of the rest seems to be handwaving. I was interested in the comment from the Yahoo! guy; technically, Yahoo! is probably the longest-standing social network of them all, so they should have a lot more useful information to contribute on the topic. As mentioned before, I've hacked up a bunch of notes on what I think is wrong - and possibly unsolveable - with social networking software, but I've yet to actually do anything like write a coherent rant^Warticle on it. Having discussed it with wisn for a bit yesterday, we both agreed that the biggest stumbling block for anyone's rating system is that the ratings are inherently at the mercies of social dynamics, as existing networks like LiveJournal have shown - hands up who hasn't had one of their "de-friended" acquaintances go off in a big hissy fit on being removed from said friendslist? Or who hasn't agonised over how someone will react to not being reciprocally friended? (or who cringes at the arbitrary conjugation of the word "friend"...)

I really should write this up into something. Although I notice Clay Shirky appears to be doing some similar exploration, except he's writing multiple articles on it.

versions

package bitlbee-0.85-1.rh90.dag (which is newer than bitlbee-0.85-1.dag.passport) is already installed
How do you tell? I guess rh90 is lexically "later" than "dag", but on the other hand it seems like the sort of call that rpm should just give up on. I guess I could rebuild it as bitlbee-0.85-1.rh90.dag.passport which I think should be newer based on what I've observed of rpm's behaviour.

public transport

( in which I whine about buses and the like )