September 30, 2004:

It's a wrap. Degunking Email, Spam, and Viruses will be on our printer's FTP site by tonight, or Monday morning at the latest. We'll have real books on or about October 22. My job is done.

Whew. Now I can catch up a little on my backlog. A week ago, at the monthly Delphi Meetup at the Panera Southgate restaurant here in the Springs, Xavier Pacheco demoed RemObjects' Chrome, a new compiler for .NET which is basically Pascal.NET. Chrome plugs into Visual Studio; if you don't have Visual Studio, you can still work from the command line. Chrome works with Mono under Linux as well, though not in any Linux-based IDE.

I just got into the Chrome beta program, but I won't have much to say about it for a couple of weeks, after I've had a chance to fool around with it some. (In the meantime, you can check out Dr. Bob's Chrome blog.) I have no experience as yet with .NET programming, and probably more valuable to me than the Chrome demo itself was a chance to get an honest opinion from some knowledgeable guys (like Xavier and Ben Oram) as to what .NET is for. I think I finally get it.

.NET is the new mainframe (erkk, dare I speak the word?) paradigm.

I used to do COBOL under TSO 20-odd years ago when I was in MIS at Xerox. The programs ran on conventional 3270 terminals, and there was nothing especially tricky about any of it. It was a reliable centrally managed code model and there was very little that the end user could do to subvert it. We used to grumble about how dull mainframe programming was, but I think we've learned in the intervening years that dull can be a virtue, especially when you have a relatively simple (if still important) job to do.

.NET is really about centrally managed code. Everything of consequence runs on the server, and the presentation layer is all that really happens on the client side. This is manna from heaven for IT geeks, who have been hoping and praying for a "PC without holes" that won't take a lot of time to support and resists end-user subversion. They've been yanking floppy and CD drives and squirting quick-set epoxy into USB ports to keep viruses out and corporate data in, and the eldest among them are probably longing for those good old (dull) 3270 days.

Xavier was pretty emphatic that .NET will never replace Win32 as the framework for standalone Windows apps, which is good. I cringe at the notion of using HTML as the presentation later of strictly desktop apps. (Think POPFile, uggh.) Still, as much of a microcomputer guy as I've always been, I understand the need for the functional model that IBM so brilliantly ruled until the late 1980s. That's where all the paying programming jobs will be in the future. If that's where you want to be, .NET is a must-know.

September 29, 2004:

We woke up this morning to brilliant sunshine, but there was fog completely hugging the ground down across Ft. Carson and the city. We're about 400 feet higher than downtown Colorado Springs, and so we were hundreds of feet above the fog, which rolled away toward the east to what seemed like infinity. It looked we were somewhere on a mountain above the clouds...because that's pretty much what we were. (We needn't emphasize that the clouds were down at ground level...)

I tried to take a picture of it, but with the sun rising in the direction the camera was looking, it was hard to get the exposure correct on my automatic camera. I was reminded of the scene in C. S. Lewis' That Hideous Strength, where Jane Studdock rises out of the fog on the road to St. Anne's, the stronghold of Logres, with the fog all around and below her, stretching away across the Midlands as far as the eye could see.

September 28, 2004:

Done, done, done. Almost. Well, the writing is done. Books being what they are, I now have to read the printed page proofs carefully, spotting typos and bad breaks and other layout entropy. That's comfortable work but it will take a couple of days. And because I also work for the publisher, I have promotional things to plan, and loads of other things.

The book should be off-press by October 20.

A huge quantity of stuff has piled up in the last month or so, but with Degunking Email, Spam, and Viruses out of my hair, I can get my workshop finished up and maybe even build something. I have some Web site stuff to finish, books to read (I'm done writing them for a few months!) and a class reunion to work on.

September 25, 2004:

Funny how trying to get a question answered online takes you down so many unexpected twisty little passages. An issue came up while preparing my September 23, 2004 entry about the Regency TR-1 transistor radio. Some articles claimed it used point-contact transistors, while most either didn't say what kind of transistor or explicitly said germanium junction transistors. (Issue like that matter to guys like me!) I realized I knew next to nothing about point-contact transistors, so I dove in and tried to learn a little.

First of all, I learned that the TR-1 used germanium junction transistors. No real surprise there, especially after learning more about point-contact transistors. These were manufactured from 1951 to about 1960, and were used almost exclusively in military and telephone switching applications. Here's one nice article on the history of the transistor that has a few pictures. Interestingly, point-contact transistors all had little holes in their cases, through which an extremely small probe could be inserted to adjust the positions of the near-microscopic cats' whisker contacts inside. No wonder the damned things cost $100 (1950 dollars, too!) apiece.

I also learned that TI created the first integrated circuit in 1958, under the leadership of TI scientist Jack Kilby. (I knew Kilby was the creator, but was not aware that the invention dated back to the 1950s.) The device, a phase-shift oscillator, is shown at left. Robert Noyce came up with a similar idea just a few months later, and he is often given credit as a co-inventor. Kilby put the first IC together almost as a stunt, over TI's summer break when almost everybody else there was on vacation. Nothing like coming back from the Catskills to find that the universe has changed, no?

Several sites (and a note from Michael Covington) taught me that the 22.5V battery used in the TR-1 was the smallest battery in the early 1950s that was not a 1.5V penlight cell, and had been used for some years in early hearing aids incorporating subminiature tubes. Most of that voltage was wasted in the TR-1, as transistors work very well on as little as 3V, especially if they don't have to generate a lot of audio power.

Those hearing aids were really the first consumer application of transistors that made significant money. The TR-1 was a stunt, actually; TI wanted bragging rights for "the first transistor radio" and sold the transistors to Regency well below their cost. The hearing aids, on the other hand, were prescription gadgetry and cost $200 or more. The TR-1's case was modeled after the pocket-sized case of the early hearing aids, which were considered viable "pocket" devices. Now that I think of it, old Louie the Barber down at his shop at Canfield and Talcott, who cut my hair over a ten or fifteen year period when I was living at home, always had a hearing aid in his white barber's shirt pocket that looked a lot like a transistor radio. As always, we build on the shoulders (and from the pockets) of giants.

I had almost forgotten about subminiature tubes, which had their brief moment in the sun in the very early 1950s. Created for proximity fuses during WWII, they made their way into portable radios for a few years until transistors drove them into history. I have a couple of them in a box downstairs "somewhere" (aye, there's the rub!) and it would be interesting to gen up a simple radio with them. (Another historical link from the Smithsonian.) The critical path would be the sockets; apparently, that 22.5V battery (which is almost precisely the same size as the modern 9V battery) is still in production, and can be had at Antique Electronic Supply for $6.26.

September 24, 2004:

In the last week or two, I've been getting clobbered by spam for a brand of coffee named Gevalia. I'm not much of a coffee freak (I have one cup of Taster's Choice Hazelnut Instant every morning to jumpstart my middle-aged synapses and that's it) and so I'd never heard of it, and wanted to check if it was a scam or something. Some research showed it to be a Swedish coffee that's well thought of among coffee addicts here, but apparently not in Sweden. I provide the link at left as an example of unintentional humor. The writer is slamming Gevalia coffee as rotgut, but note the Google-selected ads in the ad bar on the right edge of the blog.

September 23, 2004:

The October Nuts & Volts reminded me that next month will be the 50th anniversary of the invention of the transistor radio. Most people think that the Japanese invented the transistor radio, but not so: It was an American company, Regency Electronics, that got the first model into production in October 1954. It wasn't much of a radio—Consumer Reports panned it, heh—but nothing that small or light had ever been created with vacuum tubes, and even at a $50 price point (which is $329 in current dollars, according to the Inflation Calculator) it was a roaring success. The Nuts & Volts article is not online, but you can get a sense for the Regency TR-1 here.

Some facts about the Regency TR-1:

• It used point-contact germanium transistors (made by Texas Instruments) and was probably the only consumer electronics device ever to do so. Point-contact transistors were the first type of transistors made, and were fragile and more expensive than the later junction transistors, which didn't appear in quantity until early 1956.
• It used a 22.5V battery, the high voltage of which (if I understand it correctly) was demanded by the idiosyncracies of the point-contact transistors. The familiar 9V "transistor radio battery" was in fact invented by the Japanese a few years later.
• The TR-1 was widely private-labeled by wristwatch companies like Bulova, which were good at marketing high-priced small things to affluent people.

The article also taught me something I didn't know: That 2-transistor radios were created by the Japanese not because transistors were expensive, but because US tariff definitions in the 1950s considered any radio with fewer than three active devices a "toy," which enjoyed much lower tariffs. I received a 2-transistor radio for Christmas 1963, and it was crude, but so was the rock-n-roll it received from WLS and WCFL (which I had only begun to listen to that summer) and it served its purpose well.

I didn't know it at the time, but the circuit used in those 2-transistor "boys' radios" (as they were called) was diabolically clever. It was called a "reflex receiver" and both transistors were made to amplify at both audio and radio frequencies by clever combination and separation of signals on either side of the transistor. They were not superhets but TRFs, sometimes with a little carefully controlled RF regeneration. They were worthless for anything but strong local signals, but they were cheap, and (more to the point) helped the Japanese dominate the transistor radio market entirely by the mid-1960s.

Reading about reflex receivers made me wonder if I could cobble a reflex circuit up with tubes. One more project added to the list—but I still have one more chapter to finish before I can allow myself any shop-time!

September 22, 2004:

Some time back I bought a $30 package at CompUSA, giving me a PCI board containing several USB 2.0 ports, plus a small-format drive bay front plate with four USB connectors on it. I'm in the process of converting all my ZIP cartridges to 256MB Cruzer Mini Flash drives from SanDisk, and I really want a drive bay plate allowing me to plug in at least three of them at once. As you can see in the photo here, the USB connectors are so close together that you can't plug two Cruzers in side by side. Two standard USB cables can plug in adjacent to one another, but they're so close that their edges are touching. Given the amount of horizontal space available on the plate, the design is completely ridiculous.

September 20, 2004:

One of my friends reported a modest (3.7) earthquake near where she lives in eastern Kentucky, of all places, and sent me a link to an interesting USGS site that might be considered the Weather Channel for earthquakes. You can click on one of those little squares on the map to get more information, but you have to click several times to enlarge the map before the drilldown will take you to any hard data.

I wasn't aware that there was a 3.5 quake in mid-Colorado yesterday. Nothing on the news, but then again, that's barely a dish-rattler. Compare the bulk of the country to California. Blue squares mean quakes within the current day, and red within the current hour. The map is updated each time a significant quake is verified, or every hour if nothing new happens. The map is a wonderful indicator at how often measurable quakes actually happen, which seems to be, well, constantly.

September 19, 2004:

Tired, tired. Making progress, though, and I'm only a chapter and change from being done. Should finish the "change" tomorrow, and the remaining chapter by the end of the week. (Let us pray.) In the meantime, some odd lots while I have a moment here before bed:

• As reported on Slashdot, O'Reilly will begin publishing a "mook" called Make, early in 2005. A mook is a concept popular in Japan but still pretty scarce here. It's somewhere between a magazine and a book, hence the name. It's the size of a shortish book but has advertisers and some of the editorial structure of a magazine. Make will be for tech tinkerers, and that's a fine, fine thing. Dare we hope that the spirit of Popular Electronics will rise again? If anybody can make it work, O'Reilly can.
• Several people pointed out that Jude Law (see my entry for September 17, 2004) played the Gigolo Joe robot character in AI. Wow. I saw that, and didn't recognize him at all. On the other hand, I was so utterly creeped out by that film that I couldn't bring myself to review it here, and don't recall very much of it after two or three years.
• The Wall Street Journal published a short piece a few days ago reporting on some medical research indicating that premature and physically small or scrawny babies have a much higher risk of diabetes and heart and liver problems later in life (especially after 50) than babies who were born at normal weight or heavier. Here's one link. And another. The fashionable advice to women as to whether to gain weight or not during pregnancy has swing to one side or another over time, but this suggests that you might be better off putting it on during pregnancy and then working it off afterwards. I booted into the Earth program at seven pounds nought, all the more remarkable, considering that my mother was 4'10" tall and weighed about 95 pounds.
• All the chatter I've been seeing about those possibly forged Bush papers centers on the superscript "th" in "111th"—but in the September 11, 2004 entry of his Web diary, Michael Covington points out a great whack-to-the-side-of-the-head insight: The address at the top of the letter is centered, and centered accurately, which is a tough thing to do on a typewriter. Although high-end, big-bux machines existed back then that could easily do the job (I recall gazing at an IBM Composer system in covetous awe when I was in college) the National Guard would not be using them for casual memos. Michael's analysis is the best I've seen so far, and if the issue interests you, his entry on it technical, nonpartisan, and well worth a read.

September 17, 2004:

Just got back from seeing the opener of Sky Captain and the World of Tomorrow, a film for which I've been seeing previews for almost a year. It's an evocation of the sepia-hued film noir of the 1930s, with a dashing mercenary fighter pilot working from his island base in the mountains just outside Manhattan (huh? Oh yeah, this is a comic book!) that suggested a cross between Captain Nemo's Volcania and the headquarters of Doc Savage. Oddly, for all that it's set in 1939, the Nazis are nowhere in sight in this film, which might be a good thing. (We probably need to give them a rest.) The bad guy is in fact a German, but a WWI German, who has created a weird, art-deco steampunk kind of automatic island full of unlikely gadgetry. He sends flying robots and flapping metal flying wings against Manhattan, and Sky Captain (Jude Law? Who the hell is he?) flies around the globe to Nepal in his tiger-mouthed, submarining fighter plane to confront the arch-villain, with Gwyneth Paltrow in the back seat being a nuisance.

What sets the film apart is not the idea so much (which draws on the Saturday afternoon movie serials of the period) as the way it was made: By bluescreening a mere handful of live actors into what amounts to a fully animated matte painting that runs through the entire film. By not having to actually build a replica of New York City (or even the smallest part of it) for trashing, they made the film for only $70M, and the studios were licking their chops at the very idea.

Well, it may be prophetic, but this initial spin on the concept was a bitter disappointment. All too often, the film looked as though somebody were projecting a cartoon on the wall behind the actors. The bulk of the footage is film noir with a vengeance: For most of the film, the action takes place in near-darkness, making the view so murky that I had to strangle a temptation to yell, "Somebody turn on the lights!" Most of the good CGI was in the last ten minutes, but the first hour and a half was gray, dark, fuzzy, and sometimes painful to watch.

September 16, 2004:

Still pouring most of my energy into finishing Degunking Email, Spam, and Viruses, so if my entries here are spotty, it's not that I've abandoned the effort.

September 13, 2004:

One way to look at the spectrum from liberal to conservative is to map out the intensity of fear that honest people hold for cruelty and for chaos. (I say "honest people" because a lot of the noise we hear from both ends of the spectrum are from nasty phonies who just enjoy the feeling of hating some group or another.)

This is the gist of it: People who fear cruelty more than chaos tend toward the left; whereas people who fear chaos more than cruelty tend toward the right. At the extremes things get completely crazy: On the right, there are people who are terrified of any freedom they themselves don't need or understand, and on the left are people who go ballistic at the thought of anyone whom they value getting their poor little egos bruised. (Implicit in the extremes is the fact that the extremists define the extremes in terms of themselves and their anointed groups.)

You see this quite vividly in the churches, where the "law vs. love" debate rages without end. On the right, adherence to religion depends on adherence to some body of law first, and love can only operate within the framework of those laws. On the left, religionists tend to frame religious law in terms of seeing to the needs of others; in other words, a law that does not defer to love is invalid.

September 12, 2004:

Several people sent me this link, which is an article about a British researcher who is creating an autonomous, fly-eating robot. The flies are fed into a special fuel cell that digests their nasty little bodies and outputs electricity. The researchers are planning on using excrement as an attractant for the flies. Details are a little sparse, but one has a number of questions:

• How many flies must one digest to create a watt of power? My guess is that this robot had better be small, slow, and used primarily near alley dumpsters.
• Why excrement? Many years of personal experience suggest that flies are just as crazy about popsicles, potato salad, barbecue sauce, and Philly cheese steaks.
• For that matter, why flies? I always liked George Ewing's suggestion that smallish vehicles (robotic or piloted) could do well burning all those chunks of rubber that spin off the tires of 18-wheelers. On the other hand, if the robot is intended to work in New York, it could just close the loop and watch for piles of doggie-doo and shovel those into the fuel cell instead.

September 10, 2004:

The new and amazingly detailed Web server statistics that I've had since July or so (when I switched hosting services) can come in really handy. My site gets an amazingly consistent 300-350 visits per day, a number that hasn't changed hardly at all since I began tracking it in July. Then yesterday, something happened: I scored 1,459 visits in one day. Hmmm. Looking down at the referrers section of the stats, I found that siliconvalley.com had posted a link to my poem parody, "The Lovesong of J. Random Hacker." (It's down toward the bottom, in their "Off Topic" department.) Boom!

September 9, 2004:

One of the big surprises that bird watchers eventually discover is that hummingbirds are the Hell's Angels of the bird world. There are a lot of them in Colorado, even here at 6500 feet, where you would imagine their poor little wings have to work twice as hard to keep them from crashing.

Not so. We hung a small feeder up outside our breakfast nook window, and the hummers took all of about 20 minutes to find it. Three of them actually found it at about the same time, and they were not inclined to play nice. For several days we watched them rumble in mid-air, as one would begin to slurp down sugar water, only to be interrupted by the others, who would engage in a manic dominance dogfight that took them out over the back deck, swooping and diving like half-ounce F-14s. Even two weeks later, if more than one shows up, they fight. Queing up would be better, but it's pretty clear by now that hummingbirds are not British.

September 8, 2004:

Well, response came thick and fast to yesterday's entry: Apparently, watt dogs are not only older than I thought, but were actually commercialized in the 1950s. Roy Harvey sent me a link to the device at left, the Presto Hotdogger. The date is uncertain, but it's pre-Zip code and thus no later than 1963 or so. Up to six watt dogs in parallel will cook in only sixty seconds, which is about right. In looking around the Web for the Presto Hotdogger, I also ran across the Hot Diggity Dogger (at right) which was a toaster designed to toast hot dog buns—and the hot dogs too! (The perfect gift for your favorite cooking-challenged bachelor.)

September 7, 2004:

Many years ago (maybe the 70s, even) I recall a party at which Al "Cap'n Al" Duester described his scheme for "watt dogs": Take an AC line cord, separate the two conductors about 8" at the wire end, strip the conductors about half an inch, and stick one conductor in each end of a hot dog. Plug the line cord into the wall. As we all know, there's plenty of salt in hot dogs, so it will start to sizzle in no time, and in a minute or two, it's ready to drop on a bun and eat. (Do remember to pull the conductors out first—and ideally, pull the plug out of the wall before you do that!)

September 6, 2004:

Here's another Half-Serious Jeff Duntemann Speculation: The main reason the aliens have not come to visit us is that they can't afford the trip—because they're spending all their money on medical care.

The un-serious half of this speculation is the aliens. Y'all know I don't believe that there are aliens, but the Fermi Paradox is a favorite meditatand of mine, and seeing the direction that the curves of our medical spending are heading, it was an obvious conclusion. Sooner or later, the only thing we will be willing to spend our money on is medical care. Big science (or big anything) will be a very tough sell, especially if we keep retiring people at 65 or earlier, and yet keep them alive until they're in their 90s or more.

Making people live longer has been very much a rear-guard action so far. We're not fundamentally changing the aging process, but rather fixing things as they break, using increasingly sophisticated and expensive techniques, including novel drugs that must be taken for the rest of the patients' lives. The singlemindedness of the elderly about their government support programs is legendary—and how could it be otherwise? Absent Social Security and Medicare, most of our elderly would be living with their children, or in conditions of extreme poverty, or dead. It doesn't take a psychic to predict that as our elder population grows, their political power will steer more and more of the nation's GNP into just keeping them all alive.

I think it's reasonable to assume that a civilization that values life and sane government sufficiently to avoid self-destructing via war or environmental catastrophe will research ways to heal and extend life, and its ethical/political system will demand that these ways eventually be made available to all. As soon as the majority of such a civilization's inhabitants are old enough to require expensive medical intervention to stay alive, it becomes an irreversible feedback loop, and the civilization becomes a mechanism for sustaining the lives of those who wouldn't survive on their own. All other pursuits within the civilization will become secondary to that one.

Absent the discovery of some truly weird physics, interstellar travel is likely to be a very energy-intensive, expensive proposition. Once a civilization's technology becomes sufficiently advanced to travel between stars, medical technology will long since have made the majority of its people dependent on medical technology—and there won't be any Gross Civilizational Product left to building flying saucers.

September 5, 2004:

Many places are reporting a study released by the National Endowment of the Arts, showing that reading is in serious decline...except that they're talking about the reading of "literature," meaning novels, drama, and poetry. (It was decent of them not to exclude genre fiction, as they self-congratulatingly told us in the detailed report.) Consider the source, I guess. This is the National Endowment for the Arts, not the National Endowment for Knowledge. Most people in book publishing know that the real action these days is in nonfiction, not literature.

Most of what I read is nonfiction, particularly history and (of course) technology. I'd read more literature if literature hadn't gotten all self-indulgent and pompous in the last 100-odd years. Even SF, which was my mainstay for many years, is less and less a factor in my reading habits, because SF is less and less about story or ideas. Authors in the SF world are being pressed to emphasize "character development" even if they're not suited for that kind of writing. I'd rather have no character development, good ideas and a clever, rollicking plot (figure Larry Niven) than bad character development and little else. (This doesn't mean that everything can't be present in one seamless whole. It just doesn't happen very often.)

Poetry? Don't get me started. I read Robert Frost, Tennyson, and those unknown poems of e.e.cummings that actually display rhyme and meter, as well as cleverness and bright flashes of wisdom. (Read his complete works sometimes. Our man cummings was about way more than "in just spring" and pathological punctuation.)

I've always found it puzzling that anyone should recommend reading a play. That's like looking at food with a magnifying glass and then tossing it uneaten in the garbage disposal. Having read a lot of plays (I got my degree in English) and seen more than a few, I can say with some certainty that reading a play is either a waste of time or a violence to the play being read. Plays were meant to be seen.

September 3, 2004:

I'm getting tons of mail on the metapolitics of the middle, and I'll have to come back to that as time allows, which it doesn't today. In the meantime, here's a deep-geek quiz that says something about appearances. I did OK, getting 8 out of 10 correct—but then again, I've actually met some of these guys at conferences. (The programmers, not the serial killers.)

September 2, 2004:

Several people asked me to flesh out the notion of the "edges vs. the middle" in politics. (See my entry for August 31, 2004.) It's just a theory of mine, based on observations and conversations I've had with many people over the years. My metapolitical postulates:

• At the edges, politics is a tribal thing, heavily emotional and at times bordering on the hysteric. The point is to belong to and support your Tribe. What the tribe believes in is not nearly as important as the conviction that The Tribe Is Right. By contrast, the middle sees politics as a necessary evil, an intellectual exercise in governance that involves choosing among candidates based on what they've done in the past, and choosing among political strategies based on what is best for the nation.
• At the edges, party tribalists are idealists. Those in the middle are realists. There is no compromise at the edges. The middle sees all political process as the process of compromise. At the edges, it's A Holy War of Good Vs. Evil. In the middle, it's How Do We Make This Work?
• At the edges, it's about winning. In the middle, it's about governing.
• At the edges, all decisions are easy decisions. In the middle, all decisions are hard decisions.
• Party tribalists have no doubts. The middle is beset by doubts. (You rarely hear people who beat drums for a party fretting over "the lesser of two evils.")
• The edges love politics. The middle hates it.

I think at this point you get the idea.

I have to grin to read articles about some fringe interest group or another who are saying (generally through the media) to their party: "Unless we get what we want into the platform, we're going to sit this election out." Everybody wrings their hands and wonders: Are they posturing, or are they sincere? C'mon awready. Their leaders are posturing. The tribalist rank and file seem to understand that staying home only helps put the Opposite Evil in power. In any case, party leadership has pounded on them so hard for so many years to Get Out the Vote! that staying home would feel like an immoral act. They'll vote.

That being the case, the edges don't really matter much. They're committed to an ideology, and they have no choice but to support the party that most nearly supports that ideology. If their party moved to embrace the middle, they'd still support the party. Not voting is not an option, and there are only two choices to be made.

Hypothesis: The party that moves to the middle wins all the races, and rules the nation.