Gratifying

[da]

I wish I understood why it never rained but always poured. I mean, really understood, instead of shrugging and winging it each time. My leading hypothesis is a trickster god. (It's funny when one squirrel falls out of a tree onto a friend, but my money says it's gotta be staged when the second squirrel falls onto her a few weeks later. Do you know anybody who's been hit by two squirrels?)

Since I've taken the full-time job at the University a month ago, I've gotten a gratifying number of requests for my time to do small programming gigs (small as in 15-30 hours worth). The notable feature is that they travel in groups- three one evening last week, two this evening, both which I'd have probably taken six months ago.

I would really like to spread some of this karma around- does anybody know good perl programmers looking for small gigs? I've been directing people to the perl jobs list, but I'd like a bit better connection because these are all for friends or colleagues who I'd like to do the right thing by.

Comments

[dpolicar]

So, there's a classic little bit of cellular-automata work that gets used as an illustration of how groups can strongly force segregation even when no individual is him or herself in favor of it. Imagine a grid composed of cells, half of which are red and half green, each of which has the rule "don't be in the minority" -- that is, if of the 8 cells around you at least half are the same color as you, stay where you are, otherwise "move" (that is, flip color, though the total distribution remains 50/50).

There are stable integrated patterns, a checkerboard being the most extreme example (in which every cell is surrounded by half-like, half-unlike)... call that 1. There are stable segregated patterns, the most extreme being one-side-red, one-side-green... call that 0. Turns out that if you start out with random distributions, you reliably get to a stable pattern much closer to 0 than to 1.

Not surprising, really. When an "overcorrection" in one direction is stable but in the other direction isn't, the result is skewed in the stability-inducing direction.

I think the never-rains-but-it-pours phenomenon is a little bit like that. You don't get average distributions, you get spikes in both directions... but "a whole lot of nothing" is mostly indistinguishable from "a little bit of nothing" whereas "a whole lot of something" is very different from "a little bit of something."

I've been hit by two squirrels but they were stuffed and their trajectories were not random.

Anyway, Alex and Andrew might be interested in perl programming gigs.

[da-lj.livejournal.com]

ah, was it two squirrels, or the same one twice?

[dpolicar]

Two squirrels, but they were identical.

When Undead Squirrels Attack

[da-lj.livejournal.com]

ah, i was having a mental image of a furious battle in a taxidermy shop.