mult.ifario.us

I haven't made many (any?) political statements here, although it should be easy enough to guess my politics (Reed + Berkeley ≠ Republican), but I've been getting more and more fed up with what I see and hear. (Most recently, I heard on NPR that Jr. will be reading a eulogy for President Ford; I imagine that a very nice eulogy can be written for third-grade level. "See Gerald. See Gerald lead...") Even so, it's going to take more than a few years of a moron as president to kill off our culture.

On the other hand, tainting our food supply could turn out to be the lead in our aqueducts. (I know that metaphor is less than apt, as the lead theory doesn't mesh with the fact that the aqueducts ran continuously; nonetheless...) In the spirit of "what you don't know can't hurt you", the FDA declared that meat and milk from clones is safe to eat. From the executive summary:

Applied to the safety of edible products derived from clones, a finding of "no additional risk" would mean that food products derived from animal clones or their progeny would not pose any additional risk relative to corresponding products from conventional animals, or that they are as safe as foods that we eat every day. As with all risk assessments, some uncertainty is inherent either in the approach we have used or in the data themselves.

The reasoning in the report is that if a cloned animal lives to a late juvenile or adult stage, then there probably isn't anything wrong with it, i.e., damaged animals are likely to be less viable individuals. That equivalence is tenuous, but the unspoken assertion that a cloned animal is cow-equivalent for human consumption because it is viable is unfounded and probably wrong. This brings up an interesting moral paradox. On the one hand, it's immoral to experiment on people, so no carefully controlled studies of cloned meat's effect on human subjects can be done prior to introducing it to market. However, introducing cloned meat to the market is effectively an experiment on the public at large. If the meat is labeled as from cloned animals, then at least the public can "choose" whether or not to be experimental subjects or not. (I put quotes around "choose" because economics are likely to guide people's choices.)

I'll keep my purple cows on the bookshelf, thank you.

Jaxen, the ubiquitous Java XPath 1.0 library that has been in beta since (2002?) before it was cool to be in beta, finally got its 1.1 release, thanks in large part to the efforts of Elliotte Rusty Harold to clean up compliance issues and other bugs. Thanks, ERH!

Dan Pritchett's post about his career path reminded me of my experience with the academic job market back in 1997. I sent out seven applications and ended up with a good job at UIC, one of a few places (Cornell, University of Utah) with a good number of people who worked in areas closely related to my dissertation research; to put things in perspective, I was one of ~900 applicants for the position at UIC.

Before sending out applications, I asked my Ph.D. advisor about his experience on the job market. As he was finishing up his dissertation, his advisor asked if he'd like a job at Princeton. He said, "OK", and that was that. Of course, that was the 1960s... I also talked with my undergraduate academic advisor, and he ended up with five job offers — from three applications, i.e., two unsolicited job offers. Of course, that was the 1950s... Different decades brought different experiences; the mid-1970s were as rough as if not rougher than the late 1990s, and the 1980s were better (but not like the 1950s or 1960s). It takes a little doing to work it out, but the collective fortunes of generations of academic mathematicians boil down to how the academic organism responds to macro-level sociopolitical stimuli like the Baby Boom or the Cold War. This is just an anecdotal argument that one kind of prediction is as difficult as another.

As for planning (≠ prediction), the key is neither to be too specific in your goals nor to expect that being good at something means that people will pay handsomely for you to do it. I do have one piece of specific advice: don't use predictions to do planning; predictions, by nature, are usually wrong.

On Thursday last week (20061228), for no apparent reason, the otherwise incessant flow of comment and trackback spam stopped completely for about six hours and then started up again at the usual pace (~100 SPAM/hour). Maybe it was an echo of the Internet outage in Asia? Not that I'm complaining.

We had a good 2006, so I treated myself to an early Christmas present back in November — a 30-inch monitor as a replacement for my two 19-inch flat panels. It's not the most cost-effective way to acquire screen real estate (e.g., 3×19-inch has the same number of pixels (3×1280×1024 = 3,932,160 ~ 4,096,000 = 2560×1600) at roughly 50% of the cost), but I got tired of looking back and forth at two monitors. I use my eyes more than my neck with the single monitor versus the pair, so the ergonomics are better. The taller form factor turns out to be quite convenient, since more of whatever I'm working on fits on a single screen, and because the single screen looks cluttered with a number of windows that wouldn't look cluttered with two screens, I'm more likely to limit myself to only a single application at a time, which I've found is good for my productivity.

Via Dion (via Kirk, via Danny), the syntax

a -> foo = b -> foo

as shorthand for

a.setFoo(b.getFoo())

is on the table. I'll just be blunt and say that I don't like it. First off, it makes me think of Haskell, where <- and -> have meaning consistent with the use of arrows in mathematical notation, and I can't help but read the fragment above as "a maps to foo assigned to b maps to foo". (Yes, it also makes me think of the the C syntax dereferencing a member of a structure, ptr ->member as shorthand for (*ptr).member.) Second, what does it really save? At least in the example above, it saves two characters for each property ("->" in place of ".set" or ".get") and a total of four parentheses. Unlike oil and polar bears, we have an infinite supply of parentheses, so any arguments of efficiency or efficacy aren't going to win me over.

On the other hand, if some more visually appropriate syntax were to be added that supported indirect access, e.g.:

String foo = "bar";
a !! foo = b !! "bar"

you might get me interested, since this would provide a useful idiomatic syntax that would save me work with dynamic proxies and reflection.

I woke up this morning with my right hand stiff, sore, and roughly claw-shaped, but it took me a minute to groggily recall the cause. We drove up from visiting family in Portland (i.e., the one in Oregon) yesterday, and the drive was about as bad as it could have been: almost six hours straight of either crawling through congestion in Tacoma or blearily staring at the tail lights of other cars through driving rain. I tend to tighten my grip on the wheel if we hit standing water, and that was pretty much the whole trip. Maybe we'll take the train next time...