Spirit Sol 2

Oddly, after the incredible activity yesterday, not much happens today. I spend some time looking at images, but I spend most of the time finishing up those Perl scripts for Ashitey.

It's a frustrating problem: the code works, but much slower than it should. I could give them to Ashitey in this form, but I'm too proud. Eventually I trace the problem to a Perl module I'm using (I love having the source code), which makes some operations that should be O(1) take O(N). The problem shows up only when processing large files, which is exactly what I'm doing. Fortunately, the library has some other functions that I can use to get the results I want, so I end up using those instead.

Meanwhile, in other rooms, people are doing something with a Mars rover. I might as well be in Timbuktu. I'm frustrated to be out of the loop, but I decide that this is my own fault: if I didn't want to be distracted like this, I should have told Ashitey I wasn't going to write the scripts for him. But I can't help it: dangle an opportunity to write code in front of my face, and I'm hooked like a large-mouth bass swallowing a tasty worm. And, what the hell, Ashitey's a nice guy; I don't mind helping him out. In the end, I decide to look at this as my small sacrifice for the day.

But next time I'll say no.


[Five years later, I still can't believe I spent sol 2 doing that. What made it worse is that, after a couple of weeks, the scripts basically never got used again ... at the time.

But then, a few months back, Mark Maimone and Ashley Stroupe and I helped Mark Davis produce some animations for his excellent National Geographic special, "Five Years on Mars." We weren't producing the final animations that were aired; those were done by the amazing Dan Maas. But so that Dan could produce accurate animations, Dan and Mark needed to understand what Spirit had done when she was trapped in Tartarus and Tyrone, and we had the ability to produce rough animations to show them. And it just so happened that those scripts I'd written nearly five years earlier were exactly what we needed to stitch together the data from multiple sols and help smooth out the time gaps in the result.

I'd still do sol 2 differently if I had it to do over. But I spent four and a half years thinking I'd simply wasted that time, and it was nice to see that at least some payoff came out of it, in addition to my changed perspective.]