September 21st, 2010


Odin's Birds

Spent a reasonable amount of time tonight with an infrequent hobby - "visiting" other cities via GStreetView. There's something voyeuristic about doing it to non-American cities. Also, I'm trying to understand how I parse if a city is a nice place - condition of the roads and buildings seems to make a big difference, but if both are good things kind of blend together. Mexico is pretty good about StreetView - I didn't find much in South America or Africa that had any of that at all. I was not surprised to be reminded that I like Paris immensely as a city (and there are some other very likable French cities), but Edinburgh is so far the prettiest city I've seen through the service. Attn readers: are there any other cities you particularly like that have streetview coverage?

PostgreSQL 9.0 is out. Given the improvements in the area, I should probably finally learn how to configure replication. Also, the monitoring and EXPLAIN improvements sound useful, and there are some other nice goodies.

I wonder if there's any standard means for doing something nice for trashmen when one puts out a particularly large-or-heavy load for them.


Blunk and Myst

As part of my continuing microhobby of asking interesting(?) questions on AskABiologist, a recent query I made on a Soviet experiment on Artificial Selection pointed me at a very significant advance in genetic engineering four months ago that I somehow missed. Using established techniques (on the small scale) that I learned about when I took a BioInformatics class here a few years ago, they took digitised DNA and built an entire sequence to spec, which they then dropped into a real cell (replacing its original genetics), which was viable. (Their process involved more trial-and-error, of course). It's not a feat of science so much as of engineering, but it remains an impressive feat - to the extent that suitable host cells can be found (or eventually engineered), it means that we will have the ability to resurrect species (provided we collect digital DNA samples to last the meantime) after losing all physical traces of them - digital is now enough.

With the same provisions, it clears one of the smaller hurdles towards large-scale manipulation of DNA (figuring out how to do that sensibly will be a much greater feat).

It's interesting how little coverage this kind of thing gets. Maybe people don't want to think about it?

There are times when I think I would rather go into BioInformatics than CogPsy. Perhaps if no grad schools take the worm, I'll apply for those programs too. It's hard to decide which to be more excited about - the secrets of the brain or the secrets of life.