JOSHUA HOLBROOK github : twitter

AGU, Day 2 (posted 14 Dec 2010)

##I Did Something Stupid This Morning

I gave a homeless dude a dollar.

In my defense, I didn’t know he was homeless at first. He talked to me like a (mostly) normal person before we caught the same BART train, and made small talk on the way. Then, he said that he was worried about being short on bus fare. I obviously shouldn’t have been suckered in by this at all. However, if this happened in Fairbanks it would have a much greater chance of being legit. In fact, it did happen to me once. This guy, who was obviously not homeless, had forgotten his wallet, and it was cold out. He was really nice, didn’t expect anything, and when I did happen to have change he thanked me and everything. If I was suckered by this guy, at least it wasn’t too terrible an experience.

So, that’s what I was thinking when I gave this guy the time of day.

When he started saying, “five to stay alive,” any red flags that hadn’t been triggered should’ve gone up, but I guess I was just being slow. So, figuring that, if he were truly short on bus fare, that a dollar would probably get him going, I gave him a dollar.

That’s when he started getting pushy. “Naw man gimme the five!”

“I don’t have a five!” I said. This was probably trpue. Luckily for me, a Good Samaritan saved the day by telling the guy to piss off for me (for which I am really grateful). Later, another guy informed me that I “shouldn’t give them money—It makes them bolder.”

A little late. Oh well.

I Forgot About Proprietary Attitudes in Academia Again

I had a guy remind me of this while taking pictures of his poster. Nothing bad happened or anything—he just reminded me that he was planning to publish, and then he explained his (pretty cool) poster to me. But, to be honest, I did forget.

In a sense, I make a terrible academic. I like researching and everything, but I always forget to take into account this sort of thing. As a native Netizen, I have this in-built base assumption that anything being shown to me is meant to be shared. This is very different from academia, where the base assumption is that it’s a secret between you and the presenter. That is, the assumption is that you’re peer review and are secretly previewing the material so that they can be prepared to meet any common criticisms.

For me, at least, it’s weird.

Anyways: As a consequence, I’m taking down the Flickr pictures of the posters from yesterday. Sorry!


There were a lot of good posters today. In fact, I can’t really say I had a favorite poster (or posters) today! However, a few hilights:

  1. There were a few posters on predicting the level of plate techtonics on “Super-Earth” exoplanets. I talked to one guy that developed a heat transfer model based on subsurface material properties to predict the nature of subsurface convection (which drives plate techtonics) and predicted that the likelyhood was a weak function of planet size at best. This is in contrast to two other studies, one of which predicted a direct relationship and one which predicted an inverse relationship. I’d be interested to see what answer wins out.

  2. According to one poster, some earthquakes are triggered by rapid changes in the local water table! Apparently, these changes create pressure waves which can interact with pre-stressed plates to cause earthquakes. They can be pretty significant too, as it turns out, and the changes that can trigger them can be caused by both natural things like hurricanes and ahtropogenic things like dams.

So, here’s my question: The Susitna Dam would be in a location very close to a major faultline. What kind of earthquakes would that trigger? One fellow attendee suggested we build the dam as an experiment, to find out. :D

  1. There’s a crackpot poster, which I still need to get a picture of. It was amazing. I’m surprised he got poster space, to be honest.


I went and checked out the Exhibition stuff in Moscone West today, and I got a metric shit-ton of swag, no joke. I have four new posters, and a bunch of literature—some of it worthwhile, some of it not so much. I also got to talk to the people behind Alvin the submersible. Some of you may remember Alvin and Jason from the Titanic expeditions (I know I read the shit out of everything Titanic-related), but Alvin has done lots of really good science over the last 40 or so years that the program has been going. Everyone can be jealous of me starting now. :D

There were a lot of instrumentation people there. This, of course, made me think of my own forays into measurements, as well as APIs for getting the data. The best unit I saw had a web interface, which you could use to download CSV logs. What I think would be awesome, though, is a streaming data API. I think a lot of scientists would prefer to get their data pre-compiled, but even if streaming data would be unpopular amongst scientists, I think people more interested in real-time monitoring would love the streaming APIs. At least one of the vendors agreed that it was a good idea.

Expansion on Informatics Ideas from Last Post

I talked to James about this last night, and he had some really good input on the “how U data lol?” problem. Basically:

  1. Trying to make an overly-general database system is a bad idea. The better-defined your goals are, and the narrower-scope you take on, the easier it is to build a coherent system.
  2. ORMs suck, and there isn’t much you can do about it.
  3. A lot of data should just be held in its original format, on-disk.

As a consequence, we think the best system is just to have the datafiles on-disk, and use a database like sql or couchDB to handle tagging and metadata. In fact, many of the repos have the right idea.