Tablet UML News


News and commentary (and whatever else catches my eye)
from Martin L. Shoemaker, author of Tablet UML
and UML and Tablet PC instructor for The Richard Hale Shaw Group

Wednesday, October 18, 2006

Today's theory
Velcro belt straps are a conspiracy by cell phone manufacturers to sell lots of replacement phones.

Ditto belt clips that easily slide onto the belt... and easily slide off. (For novelty, some of them break off.)

Ditto tiny phones that easily fit into a pocket... and easily slip out.

And rule one of the conspiracy is simple: never, ever, ever, ever, ever offer a simple closed loop that you can thread your belt through.

Wednesday, October 11, 2006

The color Center
Alan Stewart Carl writes that There Is No Vital Center. Dean Esmay links and expands upon this, concluding:


In short: centrism doesn't really exist as a philosophy, but it's still vital. ;-)


In other words, it's the color purple: it isn't there, even though we see it.

Now to you physicists and other science geeks, this will be old news; but bear with me as I explain some color science to the audience.

Light has a frequency and a wavelength, which are interrelated through the speed of light, c. If I know the frequency, I know the wavelength, and vice versa. So we only need to look at one. Most common is to look at wavelength, commonly measured in nanometers (nm), or 10 to the power of -9 meters. And when it comes to visible light, the wavelength tells you a color. So for instance, light at 700 nm is red, light at 546.1 nm is green, and light at 435.8 nm is blue. And then there are shades in between: light at 570 nm — a point between red and green — is yellow, while light at 490 nm is cyan.

But there's absolutely no wavelength of light that is purple.

Now all of you wearing purple socks right now are looking at your feet kinda funny and wondering, "What is Martin talking about? I can see those purple socks!" Yes; but what you're seeing, doesn't really exist. To understand why, we have to look at what color light really is, and at how we see color.

First, what color is light, really? Well, with certain rare exceptions — i.e., laser beams and slivers of a rainbow — the answer is "lots". Normal light, from the sun or a light bulb or a flame or whatever — actually is made up of a spread of colors across the spectrum. If you see a blue car, most of the light you see reflected from it is blue; but there's some small amount of green and yellow and red and orange, usually; and even if there's only blue, there's a wide range of shades of blue. Unless something acts to spread the wavelengths out (like a rainbow or a prism) or selectively generate only one wavelength (like a laser), you never get just one color of light. You get a range.

Now how do we see color? Well, that's something I spent a lot of time on in my misspent youth. In an exceedingly simplified version, the cone cells in your eye — assuming you have normal color vision — come in three different varieties. One is most sensitive to red light, one to green, and one to blue. But notice that I said "most" sensitive to, not "only" sensitive to. Much like a given light spreads across the spectrum, so a given type of cone cell responds very strongly at one wavelength and then less and less strong as you get away from that wavelength. So if we go back to yellow at 570 nm, both the red cones and the green cones will respond strongly to that light. Either type would respond more strongly if we were at its "target" color; but since yellow is kinda in the middle, both respond pretty strongly, making yellow a "bright" color that really stands out. And sure enough, when you mix red and green light, you get yellow light. (If that's not the way you learned color mixing, let me give you the short answer: colors of light don't mix the same way colors of paint do. More later, but only if you insist.)

And in a similar way, cyan strongly affects both green and blue cones, because it's in between those colors. Blue and green make cyan.

So what's half way between red (700 nm) and blue (435.8 nm)? A little quick math says roughly 568 nm, or a slightly greenish yellow. So red and blue make yellow? Or green? Or... No. Just no. Whether we're talking paint or light, everyone knows that red and blue make purple — that color I said doesn't exist.

But wait! That's not what I said. I said "But there's absolutely no wavelength of light that is purple." And there isn't. But remember that most light isn't a single wavelength, but rather a blend of many, many wavelengths across the spectrum. If the light is spread across red and green and blue more or less evenly, we see white. White just means "All cones firing." But if we keep the light in the red and blue regions while getting rid of the green, we have red cones and blue cones firing, but no green. And we call that combination purple. Purple means "red and blue cones firing, but green's not involved."

So we can have purple light, but we can't have a purple laser. There's no wavelength that can be purple.

And remember, too, that red cones react to more than just red. They have a weaker response to green, and a very weak response to blue. Vice versa, blue cones have a weak response to red.

So what does all this have to do with Alan Stewart Carl and Dean Esmay and centrism? Well, it ties most directly to Dean's observation:


But after 25 years of watching politics I'm convinced that Alan is correct: there is no easily defined "center," and defining yourself as a "centrist" has no clear meaning. And I can point to two key issues to define why:

You can be against the Iraq war and against gay marriage.

You can be for the Iraq war and for gay marriage.

Name dozens of other issues, and it runs the same way: abortion, drug policy, tax policy, education policy, and so on: there is no easily-defined "vital center" because self-described centrists will soon discover that they are often at odds with each other.


In other words, if you claim to be purple (center), what you're really saying is that you're blue on some issues and red on others. And yet someone else may be just as purple, and yet be blue where you're red and red where you're blue. You may both look purple at the gross level of opinion polls. It may seem like you're both saying, "I'm independent. No red or blue for me! A pox on both their colors!" But when you come right down to it, you and the centrist next to you may disagree with each other more than you disagree with either the red or the blue. Third party proponents keep trying to find the center. They keep insisting that they and they alone have found the wavelength for true purple, without realizing that there is no such thing. There's just an overlapping range of reds and blues, with people in the middle leaning toward one color or another depending on the issue.

Monday, October 9, 2006

The world is a very strange place
There's a reason I do software, as opposed to other technical professions: it's deterministic. I can make sense of it. If there's a memory leak, then there's always a memory leak, and it's my fault that it's there. And if the leak comes and goes, it always comes and goes; and when I can figure out the pattern, I'll know where the leak is. There's always a pattern, and it's always deterministic.

When you get into the physical sciences, however, things get fuzzy; and when you get into wetware (chemistry and biology), things get downright slippery. Yes, there's a determinism of a sort at work; but the number of ways in which things can go wrong is just huge. Frankly, if we didn't prove otherwise by existing, I would say we are impossible.

Yet we exist, despite all that could go wrong; and sometimes, despite it actually going wrong, we manage to cobble together a workable life.

While researching calico cats, I came across this article on chimerism, which sounds like the inspiration for one of Stephen King's creepier works. And that article led me to this one, about a woman who learned that the children she gave birth to were not hers. Instead, they are children of her fraternal twin sister; only she never knew she has a fraternal twin sister.

See, the fraternal twin sister is her. Some of her is made up of DNA from one fertilized egg; but some of her is from another egg. Early during her development, the cells from the two embryos came together, rather than developing separately. And while this could've been disastrous — probably happens more often than we know, and probably is disastrous in most cases, leading to a spontaneous abortion — in her case, the two sets of cells were able to work things out. They shared the responsibilities of making, well, her. The cells that make up her skin and hair came from one embryo, while the cells that make up her uterus and ovaries came from the other.

So DNA testing that compared her skin and hair to that of her children "proved" that they couldn't be her kids. The two different parts of her body are no more related than any two siblings (roughly 25% common DNA).

And she's not alone. 30 such cases are known. And the chimerism may be more rampant than that:


In fact, some researchers now think that most of us, if not all, are chimeras of one kind or another. Far from being pure-bred individuals composed of a single genetic cell line, our bodies are cellular mongrels, teeming with cells from our mothers, maybe even from grandparents and siblings.


It seems that the placental barrier isn't perfect, and that some small number of cells cross from the mother to the fetus, and vice versa; and that these fetal cells in the mother can survive for decades, and might even propagate to later siblings.

All too weird for me. All too non-deterministic. Give me the simple world of binary values and logic and design. I can make that work. But this wetware has too many variables for my poor head to keep straight.

The world is a very strange place...
Posted in Opinion by Martin L. Shoemaker on Monday October 9, 2006 at 1:12am. 0 Comments 0 Trackbacks