From the Simpsons episode "Homer's Enemy" (ep. 4F19, 1997):
[Grimes walks into the lunchroom]
Grimes: Can you believe that guy? He's in his office making a pathetic attempt to look professional.
Carl: Hey, whaddya got against Homer anyway?
Grimes: Are you kidding? Does this whole plant have some disease where it can't see that he's an idiot? [Walks to bulletin board] Look here, accidents have doubled every year since he became safety inspector. And meltdowns have tripled. Has he been fired? No. Has he been disciplined? No, no.
Lenny: Eh, everybody makes mistakes. That's why they put erasers on pencils.
Carl: Yeah, Homer's okay. Give him a break.
Grimes: No! Homer is not okay! And I want everyone in this plant to realize it. I would die a happy man if I could prove to you that Homer Simpson has the intelligence of a six-year-old.
[Lenny ignores Grimes, and turns to Carl]
Lenny: So, how you doin'?
The Grimes storyline in this episode works well as an allegory for the media's interaction with the Bush administration. In this case, Homer is Bush, Lenny and Carl are the mainstream media, and Grimes is the public. Well, the left-of-center public, anyway. One could personify it in Stephen Colbert, I suppose.
Well, after having my laptop for four years, I've finally gotten it to do suspend-to-disk. I probably could have gotten that to work earlier if I'd been more bleeding-edge, but somehow APM decided to go hide in a corner while ACPI took over after a kernel upgrade. I was disappointed that there weren't any good scripts that came with the ACPI daemon, so I had to write one myself. I'll put it up here just in case anyone needs it.
I have the system suspend when pressing the suspend button (Fn+Esc), hibernating when I press the power button, and either suspending or blanking the screen on lid close, depending on if the laptop is plugged in or not. Of course, that's basically the same behavior I had before.
It's kind of annoying that some of the keypresses I used to use don't work anymore (like Fn+D for sleeping the display, or Fn+H for spinning down the hard drive), but ACPI is supposed to be the wave of the future. Well, it's been the wave of the last five years for Windows folks...
Of course, I suspect this will all be broken the next time I upgrade the kernel.
The drop of Bush's approval below 30% could be more significant than it first appears because of the way the Harris poll is conducted. Participants are asked to rate the president in one of four categories: excellent, pretty good, only fair, and poor, and undecideds either don't count or don't exist. Previous polls putting Bush at 31% did have a few percentage points of undecideds, so if you toss out the undecideds and recalculate, the 31% approval actually represents something slightly higher. The CBS/New York Times of May 5th–8th gave Bush an approval rating of 31% and a disapproval rating of 63%. The Harris poll, where there is no gap, has an approval of 29% and a disapproval of 71%.
So if you do a simple ratio (approve-disapprove)/(approve+disapprove) on the CBS/NYT poll, you'll get a value of -0.34. That's over the domain [-1,1] rather than [0,100%], so add 1 and divide by 2 to get 33%, meaning that disapproval is 67% in that poll (well, with this simple method of interpolation). So, from CBS/NYT to Harris, there is a shift of four points rather than just two. Of course, it could be slightly more likely for someone to say that the president is doing "only fair" rather than stating outright disapproval, so maybe the shift doesn't really exist.
Polls are very flaky and I don't like the way they work, especially with these giant gaps of undecideds that appear sometimes. For instance, 19% of respondents thought that Barack Obama should not run in the 2008 election, according to one poll. But another 19% felt the opposite—that he should run. A whopping 62% of people in that poll had no opinion, so saying that "only 19% of the country wants Barack Obama to run" would be entirely accurate, but also highly misleading since it would give the impression that 81% wouldn't want him to run when that's not the case.
I'd personally like a scale more like what I used earlier, where a value ranges from -1 to 1 (or -100% to 100%). A little ± attached with a value would also be nice. So, if you get a value that says 0.01±0.06, you'd know right away that it's impossible to say who's winning. This is a number you'd get by recalculating a 49% vs. 48% poll with a sampling error of 3%.
But heck, I'm not a statistician, so there's probably something pretty bogus here. Yeah, it's strange to guess at what that undecided few is thinking, but polling is already based on the fact that a small number of people can represent what a much larger number is thinking.
Today is the fishing opener, according to my calendar. No wonder it's cold and overcast.
I went to see Gomez play at the Quest last night. Very loud. The opening act was interesting because he used a sampler to loop and layer music. So, by the end of a song, you had a guy effectively playing four instruments with another four layers of vocals, though he's just standing with the microphone at the time.
Gomez played for a long time, I think it added up to two hours with the (first?) encore. I was really annoyed at first because it took about 45 minutes to get from the opening act to the main show, but I suppose I can't complain with that big of a set.
Wow, the group is ten years old. I had no idea...
I suppose I should mention that I've started to get into photography a bit more. I bought a new camera last week, and though the weather hasn't been very cooperative, I've taken a few nice shots. This streetcar photo is probably my favorite so far, though it's overexposed a bit.
My wallet didn't really like my purchase, but I know that I'll be able to make some really nice pictures with it. I don't have a great artsy eye, but I'll probably get at least a few photos to work out by accident ;-)