Michael A. Covington, Ph.D.
Senior Research Scientist
Adjunct Professor of Computer Science
Associate Director
Artificial Intelligence Center
The University of Georgia
This is a private web page not hosted or sponsored by the University.

E-mail (important messages only):
(University business only)
(all other messages)

About this notebook
Site search

Daily Notebook

Copyright 2005 Michael A. Covington. Caching by search engines is explicitly permitted.
To go to the latest entry every day, bookmark http://www.covingtoninnovations.com/michael/blog.

If you don't see what you came here for, please scroll down
(there are many topics on this page)
or look at previous months.

What Grokster doesn't grok

You've heard everyone criticizing the Supreme Court's ruling about file sharing in the Grokster case. Now listen to Charles Cooper, executive editor of CNET, defend it.

In case the link doesn't work, here are a couple of key quotes:

Justice David Souter was specific about whose feet were being held to the fire, and why. "We hold that one who distributes a device with the object of promoting its use to infringe copyright, as shown by clear expression or other affirmative steps taken to foster infringement, is liable for the resulting acts of infringement by third parties," Souter said.

I'd love to hear the would-be defenders of the cyber common weal at places like the Electronic Frontier Foundation once - just once - come out and publicly declare that Grokster's business model is predicated on breaking the law.

That is of course the point. The court decision wasn't against file sharing; it was against deploying a file sharing technology for the purpose of violating copyrights. By contrast, in the earlier Betamax case, the purpose of home video recording was simply to capture, for personal use, material that you already had a right to view. Not to redistribute it immediately to a million of your closest friends.

So the real technological tragedy here is that nobody seems to be promoting much of a legal use for peer-to-peer (P2P) file sharing. And that's a pity. It's a potentially powerful technology. It does for file storage what data packets do for the Internet, enabling it to be everywhere, free of central control. If you love freedom, you should be glad P2P was invented.

Getting back to the Grokster case, though, let me take two more jabs. First, if you look at the discussion replies to Mr. Cooper's column, you'll see that some of the pro-Grokster people believe there should be no copyrights, i.e., musicians should not get paid for their work, except incidentally.

That sounds fine if you're a rather selfish person and you're thinking about a multi-million-dollar rock star. But what about musicians who aren't rich, who are just starting out? Deprive them of copyrights, and you deprive them of their livelihood.

The purpose of copyright, as they'd know if they knew their history, is not to discourage publication but to encourage it by making it profitable. If musicians and authors couldn't make money, they wouldn't produce their works at all.

Second, Mr. Cooper takes a jab at the Electronic Frontier Foundation (EFF). I must admit I don't always see eye-to-eye with the EFF either.

This goes back partly to a computer crime that I knew something about, many years ago. The EFF swallowed the defendant's story, hook, line, and sinker, and held him up as a persecuted hero. I knew too much to believe that - but as a state employee distantly connected to the investigation, I wasn't at liberty to say so. I came away convinced that the EFF was, at best, naive.

More than that, though, I believe the EFF does not have the Internet's best interests at heart. They seem to want to keep the Internet in the hands of an aristocracy of nerds rather than deliver computing to the people.

Some computer enthusiasts want to live on a "frontier" where civilization has not yet penetrated. Why? Partly, because it's exciting, and partly, because it favors the strong and eliminates the weak. If you are strong - if you are a computer expert with lots of time to spend conquering and outwitting hazards - then frontier life will appeal to you. It will entitle you to feel superior. But if you are weak - if you are an office worker or a person with limited time and resources - then the frontier works to your disadvantage.

The goal of civilization is to protect everyone, not just the strong. And I get tired of an Internet on which I'm expected to carry my own gun, so to speak, and petty crimes against me are attempted many times per day. (I'm referring to spam, fraud, and cracking, and the expectation that I will put a lot of effort into fortifying my computer because nobody else can protect me.)

Enough of this "frontier." Let's have some civilization!

Is Windows "just DOS"?

The other day somebody tried to tell me that the Windows operating system was "just DOS inside," i.e., just a graphical shell on top of the old Disk Operating System that we used in the 1980s.

The argument was that UNIX and Linux are much more sophisticated than Windows.

Falsch. UNIX (including Linux) and Windows (NT, 2000, XP, and 2003) are operating systems in the same basic class. UNIX is better at supporting multiple users of the same computer at the same time. Windows is better at multimedia. But here are how the main characteristics of the operating systems line up:

Operating system API Memory Multitasking
DOS CPU interrupts Fixed, limited to 640K No
UNIX C function calls Virtual Yes
Windows (3.1 and up) C function calls Virtual Yes
Microsoft .NET Framework Object-oriented method calls Virtual Yes

It's true that Windows 1.0, 2.0, and 3.0 were basically shells over DOS; they got less like DOS as the operating system developed. In particular, they almost immediately got around the 640K memory limit and introduced some limited multitasking. Windows 3.1 could multitask copies of DOS. That means Windows 3.1, 15 years ago, was basically another operating system on top of DOS, not under it.

The Windows NT lineage - NT, 2000, XP, and 2003 - is not derived from DOS. It is very much the same kind of OS as UNIX (and VAX/VMS, with which it has some historical ties).

Windows 95, 98, and ME are a compromise between the two architectures, designed to be able to run DOS drivers and games. Because of this, they do have a DOS layer deep down which they can call as needed. That doesn't mean they're "just DOS." Normally, no part of DOS is running when you use them.

Even Windows XP can run DOS programs when it needs to. It does this by means of a DOS shell which you can see by going to the Run... box and typing "COMMAND.COM". Notice that this is not the usual Windows command prompt; that is CMD.EXE and is the command shell for Windows itself. The command syntax is very similar, and that has confused some people. But "C:\> dir" is not an operating system; it's just a command. You have to look inside to see how the operating system actually works.

Miscellany from the strange world of business

The work crunch continues but here are a few notes from the business world...

The recent public offering of the online gambling company PartyGaming on the London stock exchange has had amusing consequences. Basically, nobody can be sure this company's operations are legal, and there's a real risk U.S. authorities will get their act together and shut it down. So the stock prospectus consists, I am told, mostly of legal arguments, and the company's officers are in self-imposed exile in Gibraltar.

The ethics of gambling aside, here's a good argument for investing in stodgy, dull companies rather than glamorous ones.

The basic idea is that the stock price already reflects what the investing public believes about the company, and this includes all publicly available information. In the case of a glamorous stock, it also reflects over-optimism. If something is a "hot investment" then it is already overpriced - that's what "hot" implies.

Finally, here is a more optimistic view of the new personal bankruptcy laws than others I've read. I hasten to add I'm not bankrupt, but the law touches on a moral issue that interests me. Clearly, the purpose of bankruptcy is to keep people productive and to share the available assets as fairly as possible among creditors. It's not to provide a "magic eraser" to wipe out a person's just debts.

And there have indeed been people using credit cards like a casino, "cashing in" by wiping out the unsecured debts through bankruptcy. There are even (I'm told) lenders that will immediately offer you more credit cards, with exorbitant interest rates of course, right after you go bankrupt.

On the other hand, there are plenty of responsible, honest people who go bankrupt because of medical expenses, layoffs, or unsuccessful small business ventures. That is what has raised concerns.

What we have to remember is that bankruptcy is a hidden tax on all of us. Everybody else pays those bills - through higher prices at the store or the bank - when the debtor can't. More direct assistance to victims of layoffs and medical expenses might actually be cheaper.


Don't expect to see much more here for the rest of June. My day job is interfering substantially with my hobbies, including blogging. And July will also be full of hiatuses. (Note; not hiati; the word hiatus is fourth declension.)

I did want to mention one curiosity. Medical researchers report discovery of a benign virus that infects the human body, causes no illness, and kills cancer cells. 80% of the population reportedly has it. I never heard of any such thing before. Do we have normal viral flora with which we are in a symbiotic relationship?

Sigma 105/2.8 DG EX lens test

Last night, under a very hazy sky (with only first-magnitude stars visible), I was able to try out the new Sigma lens on the stars.

This is the northern part of the constellation Lyra, a stack of four 30-second exposures minus two 30-second dark frames with the Digital Rebel at ISO 400, piggybacked on my old faithful Celestron 5.

On the full-frame picture, there's appreciable vignetting at f/2.8, but it was hard to assess because the sky background was uneven. I don't consider it a crucial problem because I'll often be using only the central portion of pictures taken with this lens.

Here you see the middle 1/3 of the frame vertically and middle 1/4 horizontally. It's sharp! In fact, the whole picture is this sharp, all the way to the corners.

Note the cleanly split double star Epsilon Lyrae near the top. Those stars are 208 arc-seconds apart, or 0.1 mm apart on the CCD. You can see that the star images are less than 0.02 mm in diameter. That's what we usually call "pinpoint stars."

In fact, Zeta Lyrae, the bright star toward the bottom left, is also double, and you can see it elongated here. Before I downsampled the picture for the Web, it was almost split. Its separation is only 43 arc-seconds, or 0.02 mm on the CCD.

Mechanically, this lens seems solidly built and feels more like a Nikon than a Canon lens, even though I'm using the Canon-mount version. The infinity mark is precisely correct, which is often the case with Nikon lenses and less often, in my experience, with Canon. It's one of those little details that are handy for astrophotographers and more or less irrelevant to anyone else. It shows that either I got lucky, or the lens was given its final adjustment by a very picky optician.

Was it a lie or a trick?

A woman was promised "100 grand" as a prize by a radio station, and it turned out to be a Nestle's "100 Grand" candy bar.

She's suing them. I'm inclined to take her side.

The radio station knowingly and intentionally caused her to believe the prize was $100,000, which is how the phrase "100 grand" is normally understood in the absence of indications to the contrary.

And she was entitled to the truth. Sometimes a misleading statement is justified because the hearer isn't entitled to know the truth (e.g., to safeguard privacy). This isn't one of those cases.

Their defense is that they didn't mean what they sounded like they meant. Hmmm...

If that were so, anybody wanting to deceive the public would have carte blanche to make up their own meanings for words - or find obscure, rarely used meanings - and claim that whatever they said, it had those peculiar meanings rather than the obvious ones.

I'm a linguist. I know very well that words don't mean things, people do. They knew very well that they would be creating a false belief.

If you apply this criterion, a lot of the "tricks" used in advertising don't pass ethical muster, such as junk mail disguised as FedEx envelopes or (in the old days) Western Union mailgrams, to get your attention.

It's wrong to make people believe false things. That's what we call lying.

The source of an obnoxious noise

For some time I've thought there was a defect in my scanner radio. Every so often it would make a strange noise, whether the squelch was open or not.

Then I heard the same noise, much louder, in my computer speakers. That led me to think it was some kind of electrical noise on the power lines in my house.

Then I heard it on a friend's karaoke machine. (Note concerning karaoke with members of the Linguistics Program: Phoneticians and dialectologists sing well. I am a semanticist.) Anyhow, that convinced me it wasn't confined to my house.

Today I tracked down the source: my cell phone and presumably other cell phones too.

The sound is a galloping rhythm of buzzes, if that makes any sense. To hear it, click here. (Isn't that the ugliest thing you ever had an MP3 of?)

It is transmitted when the cell phone identifies itself to the network. Your cell phone transmits occasionally even when you're not talking on it. It has to tell the nearest tower where it is, so that incoming calls can be routed to the right tower. This happens at irregular intervals perhaps every 20 minutes. That's why they want you to turn your cell phone off when you're around sensitive equipment - not just refrain from making a call.

The interference in my scanner radio is probably unavoidable because the radio is designed to pick up very high frequencies. I'm going to try to clean up the computer speakers, probably by adding a small capacitor or a ferrite bead on the input.

Happy Father's Day to me: My devoted family gave me several presents, one of which didn't arrive until today: a Sigma 105-mm f/2.8 EX DG macro lens for the Digital Rebel. This is the latest descendant of the superb Sigma 90/2.8 macro, which I liked so much that I've bought both the Olympus and the Nikon versions.

It's an 11-element version of the venerable Zeiss Sonnar, which, I think, had 4 elements. Eleven. Can you believe it? With multicoating there will be no internal reflections to speak of.

I haven't tried the new lens very thoroughly, but what I like about the earlier version is that it is very sharp not only in extreme close-ups but also at infinity. I've done quite a bit of astrophotography with macro lenses!

The other presents took the form of a box of edible goodies, especially nuts, presumably on the principle that you are what you eat.

Tri-state TreeView in pure .NET

One of the annoying omissions in the .NET Framework TreeView control is that although the nodes can have checkboxes, the checkboxes have only two states (checked and unchecked). For some purposes you need a third state, a gray checkmark to indicate that some of the subordinate nodes are checked but others aren't. Here you see how that's done in NTBACKUP, the Windows tape backup utility (see picture). It also has other uses.

There is an excellent control called TriStateTreeView on the The Code Project. It implements exactly what I need, but it's GPLed (meaning I can't use it in commercial work unless I'm willing to share the source code) and it relies on the Win32 HitTest message to find out whether the user clicked on the text or the checkbox.

Can I make my own, without using HitTest or any other low-level Win32 calls? Indeed I can. Because it's needed for a commercial project, I can't share my code, but here's the general approach.

  • The overall program logic is similar to TriStateTreeView, which I recommend studying.

  • The TreeView has an ImageList comprising empty, checked, and gray-checked boxes. The TreeView does not have the CheckBox property turned on; I'm using images (icons) instead.

  • I handle the MouseDown event of the TreeView and determine, each time, whether the user has clicked on a checkbox icon. This is done as follows:

    • The MouseDown event arguments contain the mouse position, e.X and e.Y.

    • The TreeView has a GetNodeAt(x,y) method which identifies the node you've clicked on. If it returns null, bail out because you haven't clicked on a node.

    • Assuming you got a node, look at its Bounds property to find out the position of its text.

    • If the mouse position is to the left of Bounds.Left, by at least 2 pixels and no more than 2+W pixels where W is the width of the image, then you've clicked on the checkbox icon. If not, bail out.

  • From there, the logic of what to do to the checkboxes is the same as in TriStateTreeView. I haven't implemented all of it yet.

As you might guess, I'm going to use this to enable users to select files and directories, much the way it's done in a tape backup program. The art of putting files and directories into a TreeView is dealt with neatly here (in Visual Basic). Crucially, there's no need to populate the nodes underneath a given node until the user expands it. Thus, you don't have to search your whole disk right at the start.

Folks, that wasn't identity theft

In the wake of the CardSystems security breach, the news media are bristling with stories about how to defend yourself against identity theft.

Folks, the CardSystems breach was not "identity theft." It was fraud or theft of information inside a business that processes credit cards charges.

The thieves will be putting false charges on people's accounts. They will not be "stealing identities" or impersonating the people in any way that goes beyond use of a single account.

I think what we are seeing here is the journalistic mind at work. Journalists think in preconceived categories. They've been looking for "identity theft" and something comes along that resembles it, so they force it into that category.

As I've said before, I think the term "identity theft" is misleading in the first place. Anyone who "steals my identity" isn't stealing anything from me; he's trying to steal money from my bank. Not the same thing.

Why is this notebook suddenly getting over 200 hits a day? Did I blog something important?

Odd optical gadget of the day

Gadget of the day: A quilters' kaleidoscope, or at least that's how it's marketed:

And here's what you see when you look through it at the clock on the wall:

Not exactly a quilt, but an amusing effect.

This is going into Melody's kaleidoscope collection.

Wristwatch found

My Quixote quest for a wristwatch (see also here) has come to an end with a Casio A178W (pictured; Casio photo).

This one has a metal body and takes standard bands, so I was able to put a good Speidel band on it. Its controls are exactly like the one I was replacing. But its display is slightly less informative; there's one less numeric field, hence no date while viewing the time, no home time while viewing the second time zone, and no time of day while viewing the stopwatch.

I'm going to rig up a band for the old watch so I can wear it while doing astronomy. Because it only accommodates 10-mm pins, and no replacement for Casio's odd sports band is available, it may have to have a ladies' band, but I'm secure enough in my masculinity to live with that. Since the controls are the same on both watches, I won't be pushing the wrong buttons when I'm busy taking pictures.

Important note to Casio: I chose the A178W because of its lack of features. It has no modes or buttons that I don't actually use. I won't have to navigate through unwanted menus to get what I want. Everything is simple.

It also looks like something out of about 1985, but that's fine with me! It's designed to do its job well. The fake-1930s watches of 2005, with their multiple illegible dials, annoy me.

The American credit card industry crashes in flames...

Well, not quite. But the house of cards may be about to topple.

My Discover card, which Melody and I use for nearly all expenses and pay off monthly, was revoked suddenly yesterday afternoon in the middle of a shopping trip.

We called Customer Service and were told it was due to a security breach. Sure enough - we're one of the victims of the massive intrusion at CardSystems in Tucson.

And the way Discover reacted was to revoke our card, suddenly and without consulting us, and issue us a new one, which will arrive in a few days.

If we had been traveling, this could have been a disaster.

So now there's a new risk associated with credit cards. Any credit card can go dead suddenly because there's been a crime hundreds of miles away. Not good.

I'm sure the FBI is pursuing numerous possibilities, including the possibility that CardSystems was corrupt from its inception. I'm also sure that no accurate technical details of the intrusion have been given out, or will be given out until the case goes to trial. Suffice it to say that this is not "phishing" (impersonation of a bank) or identity fraud (large-scale impersonation of an individual). This is good old-fashioned computer crime; somebody wanted lots of credit card numbers and figured out where to get them.

But my concern is that credit cards are much less useful than they were last week. Forty million card numbers were stolen. If all those credit cards are revoked, even just for a few days, the cost to business and to individuals will be immense. Think of the purchases that weren't made because people's credit cards didn't work.

This is one more dubious development in the history of the credit card industry, which started out as a convenience for business travelers, then became a substitute for department-store installment plans, and finally a vast loan-shark industry that lends huge amounts of money and reserves the right to raise interest rates on anybody at any time.

I think it's a house of cards ready to tumble. The credit card industry could lose a lot of money in a hurry if the following things happen:

  • Fear of fraud (or fear of heavy-handed anti-fraud measures) leads people to stop using credit cards suddenly.

  • Fear of suddenly rising interest rates...ditto.

  • Fraud itself costs the industry more and more money. Remember, the bank eats the cost of fraud, or passes it along to customers.

  • A sudden rise in interest rates leaves people unable to pay their bills. Credit card loans are unsecured, and normal practice is for the bank to simply absorb debts it can't collect. Some people (who don't care much about their credit ratings) know this and use credit cards like a lottery, hoping they'll never have to pay them off.

I'd like to see legislation and/or regulation to do all of the following:

(1) Require basic anti-fraud measures including a high level of security everywhere cards are processed, and basic ID checking at every purchase to make it hard to use a stolen credit card number.

Part of the problem right now is that the system leaks. In an effort to make online and phone shopping easy, and to accommodate small purchases, the credit card industry has made it entirely too easy to use someone else's number.

(I also think it would be a good idea to poison the database. Anywhere large numbers of credit card numbers are stored, make sure at least 10% of them, randomly scattered, are booby-trapped numbers which will set off alarms if used. That would make it much harder for people to get away with stealing numbers.)

(2) Require minimum payments high enough to pay off any loan in 10 years. This, fortunately, is already happening. I just hope it will be phased in gradually.

(3) Ban loan-shark tactics such as offering low promotional interest rates and then jacking them up suddenly when the customer is late with one payment.

The essence of usury is taking advantage of people's misfortunes by charging them exorbitant interest. That's exactly what credit card companies do. And because of a dastardly provision called "universal default," being late with one payment can trigger high interest rates on all your cards.

I'd like to see a statutory maximum interest rate - maybe 15 percentage points above the federal funds rate - and a law that the interest rate on an existing loan cannot be raised more than 2 percentage points per month, regardless of the reason, and regardless of any prior notice or agreement that the customer may have supposedly accepted. If they want to take your rate up from 5% to 29% (which is presently done at the drop of a hat), they'd need a whole year to do it.

Or better yet: The interest rate on an existing loan cannot be raised at all. Every month the credit card company makes a new loan to you for your new purchases and sets the interest rate. What you already owe will keep its interest rate until paid off.

If banks can't make loans under these terms, maybe they shouldn't make them. What happened to the idea that credit limits were based on ability to pay? In the 1970s, it was common for a well-to-do person to have a $500 credit limit; today it's common for your credit limits to total a year's salary or more.

The opinion has been expressed that credit cards are better than fly-by-night loan sharks; people with bad credit are better off borrowing from a bank than from the Mafia.

But that's like saying "legalize heroin." Lending money to people who can't pay it back is wrong, period. So is making the rest of us absorb the cost of the defaults.

Are professors being corrupted by bribery?

The latest Reader's Digest, in a gossip column titled "That's Outrageous!", reports that college professors are being corrupted by "payola." Specifically, they are being bribed to:

  • send to newspapers, under their own names, political opinion pieces that are ghostwritten for them by special-interest groups;

  • require students to buy particular textbooks.

Well... "Payola" must be fairly rare, because in 23 years of teaching, I've never heard of it. I'm afraid this gossip column will get the public confused about the legitimate consulting and reviewing activities of professors.

Most college professors are not full-time 12-month employees. They are typically employed 9 months out of the year and are encouraged by their employer to do consulting for industry, including the publishing industry. This builds useful contacts and prestige, and it does not corrupt their classroom activities in any way. Academic publishing could not operate without it, because a publisher cannot maintain a gigantic staff of experts each of whom is needed only a few hours per year.

The last thing we need is someone telling the public that these legitimate activities are "payola."

Actually, state university professors are government employees who could be prosecuted if they took bribes. The people paying the bribes could be prosecuted too.

There's always gossip, among the students, that professors are being bribed to require expensive textbooks. And there's always gossip about horsemeat being served in the dining hall. But that doesn't mean either of these things actually happens. In 23 years I've never heard even one concrete accusation.

As for signing a ghost-written political opinion piece, a professor would have to be fairly stupid to do that. It's himself he's making a fool of. Any professional scholar knows that his reputation for expertise is his most important asset.

End of an era

Kodak is indeed discontinuing black-and-white photographic paper. It's the end of an era. Quite a few smaller companies are marketing paper successfully, such as Bergger and Arista. Kodak's major competitors, Ilford and Agfa, are still selling paper, though in financial difficulty. I don't know whether Fuji still makes black-and-white paper for the Japanese market.

The picture shows how Kodak paper was packaged in the mid-1960s. I made my first prints in 1969, using Medalist paper from a very similar package.

In the early 1970s, we got resin-coated (RC) paper, which made fixing and washing much faster, but triggered a long debate as to whether the stuff was really reliable. To this day, some art galleries will not take photographs ("silver gelatin prints") that are printed on RC paper, though they will gladly accept inkjet ("giclée") prints on much less proven materials.

The main technical fault in black-and-white prints 30 years ago was poor, grayish blacks. In the early 1980s, after the Hunt brothers engineered a silver shortage, the word got out (falsely) that prints had poor blacks because there was no longer enough silver in the paper. This started a quest for darker and darker blacks, as well as initiating a market that continues today for "premium" or "high-silver" or "high Dmax" fine-art photo paper.

And it was misguided. It was shown that the best prints by Ansel Adams and others had maximum blacks of about 1.8 (i.e., reflecting 1/101.8 of the light) whereas the paper was clearly capable of 2.2 or more, silver shortage or no silver shortage.

Kodak and Ilford addressed the problem by changing the characteristic curve of the paper so that it would give higher contrast in the blacks than in the whites. This is the "Polycontrast IV" and "Multigrade IV" look.

There was also uncertainty about whether the paper should have the developing agent incorporated into it. In the 1980s, many RC papers did; the purpose was to facilitate machine processing, and to eliminate one cause of "poor blacks," namely incomplete developing.

But then it was shown that a print need not have maximum development in order to have good blacks. Kodak introduced a paper, Polyprint (later Polymax), that allowed the user to vary the development to control overall density. I adopted it enthusiastically.

Meanwhile, Ilford was pioneering ways to make fixing and washing go faster. Kodak's paper has a relatively hard emulsion that requires 1 minute to fix and 4 minutes to wash. Ilford's paper goes through those steps twice as fast, which means that if you use Ilford's paper and chemicals with Kodak's times (as I do), you can have complete peace of mind about the permanence of the print. About 15 years ago, I had problems with incompletely fixed Polyprint prints and, sad to say, switched to Ilford, never to come back.

Other companies will keep making black-and-white paper indefinitely; it's not complicated stuff. But I doubt that we will see any technical improvements. This is the end of the line. I'll order one last box of Polymax (which has already been discontinued), if I can get it. The Kodak paper that sells out last will probably be Azo, one of Kodak's oldest products, a contact-printing material that has changed little since the 1920s.

Wristwatches and ink-to-information ratio

As I mentioned yesterday, I've had to retire my Casio W-96H wristwatch because, in just two years of daily use, its strap has broken, it does not take standard straps, and the print is starting to wear off the housing.

This is a pity because I really like the way this watch works. I may even drop $17 for another one just like it at K-mart. I very much regret that Casio doesn't make a more durable version.

Why do I like this watch? Partly because its stopwatch is just what I need for astronomy: the stopwatch isn't disrupted by actuating the light, and I can see the time of day while using the stopwatch.

But more importantly, I like its low ink-to-information ratio. It has a rather big LCD display, and everything on it is informative.

This watch isn't bedecked with circles and lines just to make it look complicated. It communicates concisely and gets right to the point.

The concept of ink-to-information ratio was introduced by Edward Tufte, one of the deep thinkers in contemporary graphic design. The idea is very simple: Don't put ink on the paper if it doesn't help convey the message. Mr. Tufte was thinking mostly of graphs that are criscrossed by grid lines and the like, but the same principle applies to watch faces.

By the way, part of my problem is that digital watches are strongly out of fashion right now. You're supposed to have an analog dial like a Rolex. If you wish, you can have a complicated watch with lots of little dials within the big one. What you can't have, except at the very low end of the price range, or on fancy sports watches, is digits.

I guess I'm a 1980s nerd. I see no reason why an electronic watch has to have hands.

Casio photo.

What will Mars do on August 27, 2005?

Mars won't be doing anything special on that date. But a chain letter is circulating on the Internet that says Mars will be unusually close to the earth on that date and will look as big as the full moon.

Mars did make an unusually close approach on August 27, 2003, but it only looked like a bright star. Details here (second item on the page).

Moon and Jupiter

The moon passed near Jupiter in the sky tonight. I snapped this 1/400-second exposure at ISO 800 with the Digital Rebel and an old Spiratone 300-mm f/5.6 mirror lens on a fixed tripod. I used Photoshop to brighten Jupiter separately from the rest of the image, and to sharpen the moon.

Wristwatch wanted

I'm looking for a wristwatch for everyday wear (including moderately dressy occasions) which is also suitable for timing astrophotographic exposures. Here are the attributes I want:

(1) All-digital display (no hands).

(2) A stopwatch function that is not disturbed by operating the light. Many watches use the same button to turn the light on and to reset the stopwatch or kick it into lap-timing mode.

(3) Ability to read the time of day (in a smaller display) while using the stopwatch.

(4) Durable construction, preferably with a metal body.

I have a cheap Casio watch that satisfies (1), (2), and (3), but its band has broken, it doesn't take standard replacement watch bands, and the paint is wearing off its body.

I'd like to find a well-built metal-bodied watch with the same firmware.

I also have a very interesting (and amusing) Casio watch that calculates sunrise, sunset, and the phase of the moon as well as trying to prognosticate fishing conditions (an earlier version of this one). Its stopwatch satisfies (2) but not (3). I've gotten rather poor battery life from it - a CR2016 cell every year or so. Battery replacement is supposed to be done only by Casio, but that would cost as much, each time, as buying a new watch!

A word to the wise: do not handle watch batteries (or other button cells) with metal tweezers. Doing so shorts out the cell and greatly reduces its life. It would probably only take about 30 seconds to ruin it completely.

A couple of years ago I refused to buy a watch battery because the store clerk was installing it with metal tweezers. I suppose she had gotten away with this a few times - and the watches worked for a few days, at least - but it's definitely not a good thing to do.

More miscellany

Network change: www.ai.uga.edu has successfully moved to Well, almost moved. We're still having a little trouble with our VPN and some other internal things. I worked on it 5 hours yesterday even though it's supposedly no longer my job; I thought I could save them some time since I was the one who set everything up last year, before the move.

Wedding guest blues: Read this interesting and well-informed rant about the high costs imposed on wedding guests by "Bridezillas." (I wonder how many of them are Nth marriages, for some sufficiently high value of N.)

I agree wholeheartedly that costs should be kept down, but I'm a little worried that the notion of "Bridezilla" may become a handle for disparaging marriage itself. "Don't have a wedding, it's crass, your friends will think you're greedy... Just move in with him." Falsch!

The end of online gambling? Professor Bainbridge is complaining about people who write computer programs and let them play online games, especially poker, as if they were humans.

I've seen this demonstrated 4 years ago (and sternly cautioned the implementor never to use it in a real competition). I agree, it's cheating to enter a machine in any kind of humans-only competition. It would be like surreptitiously using a bicycle in a foot race.

But unlike Professor Bainbridge, I don't approve of the current poker fad (see also this). Maybe computer players are what will kill it off, at least the online version!

I also strongly support proposed legislation to ban credit card companies from doing business with gambling sites. Moneylending and gambling don't mix.

Last, and funniest: A few weeks ago I got e-mail from someone who found the word "gambling" on my web page and decided I must want to promote his online gambling site!


By yesterday evening all the news services had retracted all of the day's news about Natalee Holloway. No confession; no admission that "something bad happened"; no news. I will not try to log further developments.

Strange tales of law and business:

(1) Illinois lawyer accidentally sues himself. It had to happen eventually... the legal equivalent of a gun accident, I suppose. Shorter account in News of the Weird.

(2) People are finally realizing that homeowners' associations can lower your property values by driving away buyers. (Follow the link and scroll down.)

At last! There have been scattered complaints about draconian restrictions around Atlanta (e.g., you must paint your house a certain color, you cannot keep a pickup truck at your house, you cannot have any kind of antenna - that last one may be subsidized by the cable TV company).

Let's put out the word: People don't like unreasonable restrictions. Homeowners' associations are not landlords.

Yesterday evening I was delighted to learn that one of my favorite pieces of music - which I knew only as a hymn, "Praise to the Lord, the Almighty" - is from a Bach cantata. And Bach's version of it is breathtaking!

To hear it, click here, scroll down to the listening samples, and choose number 14 in the "Windows Media Player" column. The original tune will not jump out at you because Bach has built so much else around it, but it's there.

This is more support for my theory that all the really good music in the world was composed by Johann Sebastian Bach. All the rest are just imitators or precursors.

Aruba tragedy / How "news" develops

Right now (11:30 a.m.) different news media are saying very different things about the mysterious disapperance of Natalee Holloway in Aruba.

[Note added June 12: Due to subsequent retractions, the following links probably no longer lead to the stories that I was commenting on. You will probably see a later version instead.]

CNN says one of the suspects has confessed to killing her.

The Associated Press only says that one of them admitted that "something bad happened" to her. Fox News says the same thing in more depth.

The BBC apparently isn't reporting any new developments at all.

Whatever it is, though, it's sad. Our prayers are with her.

It may seem crass to say so, but I think a lot of people are just now realizing that "partying" can be a very dangerous activity and that this young woman was sent into a dangerous situation by people whom she trusted. I wouldn't send an 18-year-old to drink heavily in an unfamiliar place with no one keeping track of her. My daughters (ages 16 and 20) agree wholeheartedly with my position.

(Whether Natalee was a legal adult is beside the point; adulthood does not come all at once, and nobody is ever totally self-sufficient. As best I can tell, Natalee made some bad decisions herself and was given bad guidance by other people. She went on the trip thinking it was safer than it actually was.)

Why people like to make things

Ulla-Maaria Mutanen has published this "manifesto" about the satisfactions of making things.

I was referred to it by Make Magazine, an up-and-coming new publication full of genuinely new ideas. Make's key insight is that nowadays, most electronic or mechanical projects are modifications of pre-existing gadgets, not built from scratch. We might as well exploit those cheap Far Eastern factories... They're onto something good. I'm mildly annoyed by the magazine's undercurrent of "we are a very fashionable in-group" but maybe that's fading.

Getting back to Ulla-Maria's crafters' manifesto, I would add one more thing:

Creative construction is the most uniquely human activity and gives us a taste of God's creation.

J. R. R. Tolkien developed the concept of sub-creation, the process by which a person composing a story tastes what it was like for God to make the world. I would add that engineers experience sub-creation at least as truly as fiction writers, if not more so. Machines to have to work; they have to obey their own laws. That is exactly what God was faced with when He invented the physical universe.

Carpeting Minerva

That sounds like a parody of "Waltzing Matilda," but actually, it's what I've been doing.

Minerva is my home desktop computer, a 2.4-GHz Pentium 4 in an excessively large tower case. A few years ago I bought the biggest case I could find - back in the days when you had to have a diskette drive, a ZIP drive, a tape drive, etc. - and I've been moving up from one motherboard to another for some time.

And Minerva was noisy. The big case acted like a sounding board for all the fans and disk drives.

So this week's project was to make Minerva quieter. I know the proper way to do this is to use a new mini motherboard and tiny case. Computers are getting smaller. But I didn't want to make such radical changes.

Accordingly, I went to Home Depot and got some cheap indoor-outdoor carpeting that doesn't unravel or shed lint at the edges when cut. Generous amounts of this are now stuck to the insides of Minerva with carpet tape.

I also installed a quiet power supply.

And I relocated the disk drive bay. The hard disks reside in a metal frame that happens to block access to numerous important connectors on the motherboard. This metal frame also works almost like a loudspeaker cone to amplify their whine and transmit it to the rest of the case.

I detached that metal frame from where it was mounted; added a fake disk drive made of basswood, to absorb vibrations from the real one; and attached the frame to the bottom inside of the tower case using industrial-strength Velcro. A label on the outside of the case warns anyone who might want to move the computer that the drive bay needs to be bolted in place first.

The result? Minerva is much quieter. Now I can hear the whining of the other computer in the room.

Software to increase your computer's memory?

Ever since Windows 95 I've heard about software that supposedly increases the amount of memory available on a computer. I thought it must be hogwash - what sanely designed operating system would leave some of its memory unavailable? But then, not all operating systems are sanely designed...

I finally found out what's going on, and it is hogwash. See this and especially this.

The "memory optimizers" free up RAM by kicking other programs out of it. The total of available memory goes up, of course, but it's memory that Windows would have reclaimed anyhow if it needed it. In the meantime the memory contained something that Windows decided to leave there because it was potentially useful.

So a memory optimizer makes your computer report that more memory is available, but it doesn't perform any better.

Notice that I'm talking about RAM, not disk space. Compressing files can certainly save space on the hard disk.


Soundcards are better than audio-on-the-motherboard. My Asus motherboard has a built-in SoundMax system, but when I record from Line In, there's some noise, mostly at low frequencies, about 50 to 60 dB below maximum level. That's not bad; it's comparable to a cassette tape. But when I put in a Sound Blaster Live, I was glad to see that the noise went away. The Sound Blaster's noise level is more than 95 dB below maximum. Why? Less electrical noise from nearby components, I think.

By the time you read this, you may not be able to get to www.ai.uga.edu or related sites where I work. We're changing the IP numbers of the entire network.

Just wait a few hours or days for your DNS server to pick up the new address, or click here to try the new number.

Fortunately I don't mean "we" literally. I have three very good technicians masterminding the job.

Maybe we'll finally lose the people in Europe who have been trying to break into our FTP system every few minutes for about three years solid. Suffice it to say their technique has proven ineffective, but their determination is...well...puzzling.

Useful possession acquired recently: a sleep mask padded around the edges so it doesn't press on my eyelids at all. If you keep astronomer's hours, or if you travel a lot, a sleep mask is a must. This one is expensive but worth it (to me).

How I fixed the Linksys WRE54G

[Note: Some of the problems with this device were never resolved and it was taken out of service in July. See July 21.]

I finally got my Linksys Wireless Range Expander working in the following way.

The setup utility was no help, and neither was Linksys tech support, which sent me the canned response to a completely different question than I asked.

The problem was that my network has WEP (Wireless Encryption Protocol). I'm not running a network for the whole neighborhood and any spammer who may drive by. Network Stumbler (a very handy piece of software for measuring signal levels) tells me one or two neighbors have completely unsecured networks that I could get into if I wanted to. I know WEP isn't unbreakable, but I figure nobody's going to bother to pick my lock when most of the neighborhood has no locks at all.

Anyhow, the procedure for setting up the WRE54G was the following. It was very clumsy and involved loss of connection several times:

  • Disable WEP on the main wireless access point and on the PC from which you're doing the setup.

  • Reset the WRE54G and press the Auto Configuration button. It automatically sets itself to the same channel and SSID as the main network.

  • Get into the WRE54G at and enable WEP on it, with the appropriate key.

    Problem: The WRE54G appears to hang in the middle of this process. Actually, I think what happens is that it loses connectivity with my PC as soon as it enables WEP.

  • Regain connectivity with the main network; this was awkward and involved disabling and re-enabling my wireless card.

  • Re-enable WEP on the main network, with the same key.

  • More disconnecting and reconnecting...and eventually...voilà...it works!

A couple of key points about a repeater (which is what the WRE54G is):

  • It is not like a second access point. It shares the SSID and channel of the original one; it doesn't have its own.

  • It costs you half of your bandwidth because it retransmits every packet to the main access point.
But the signal strength in our den is more than 24 dB better - that's more than a factor of 200 - and we're happy.

Odd HP LaserJet 1200 problem

No sooner did I get the WRE54G fixed than Cathy called for help. Our home server, Eos (more formally Rhododaktylos Eos because it was bought at a 6 a.m. sale), was no longer printing.

When we installed the LaserJet 1200, HP's software created a new virtual port for it with a name like DOT004 or somesuch, presumably in order to do tricks that an ordinary parallel port can't.

I switched it back to LPT1: (the normal parallel port) for each of its four drivers, and everything was OK again. I don't know what this DOT004 business is; reinstalling the HP drivers didn't fix it.

Telescopios modernos para aficionados

   El libro mío sobre Telescopios modernos para aficionados ha aparecido en español.

That is, How to Use a Computerized Telescope has been translated into Spanish and published by Akal Publishers, Madrid.

Two copies arrived unexpectedly today; I had had only a dim idea that the translation rights had been sold! I haven't peeled the wrapper off and read the book yet. Too busy fixing computers!

What to do when .chm files won't open

My laptop could not open a .chm (compiled HTML help) file that contained the documentation for a soundcard. I got a barbarous error message somewhat like this:

Cannot Open the File: Mk:@MSITStore:C:\Temp\SoundBlasterLive.chm

I had exactly the problem described here except that I'm running Windows XP, not Windows 98.

Solution: I went to a command prompt and typed:

regsvr32 hhctrl.ocx

and everything worked again.

I think that some recent Windows update has accidentally left part of the help system unregistered.


I'm not sure I've given up on the Linksys Range Expander, but I don't trust it. In general, I don't like for wireless equipment to be reconfigurable only by people connected to it wirelessly. Doesn't that open up a lot of security risks? I'd rather hook up a cable to configure the equipment, thank you. That way I know nobody else is hooked up the same way.

The Supreme Court has refused to change an unpopular law about medical use of cannabis. Personally, I think medical use of cannabis should probably be legal, though it shouldn't be used as a cover for recreational use, and smoking is not a suitable way to administer the drug. I must admit I'm unsure what medical conditions it is good for (I definitely know some it's bad for). But the drug is not so overwhemingly dangerous as to preclude any medicinal use, nor is it equivalent to some other, safer drug.

But my point is, the Supreme Court wisely (for once!) chose to be a court, not a legislature. They pointed out that Congress can change the law; they didn't change it themselves. Their job, as they saw it, was to rule on what the law is now. A small victory for government by the people.

Short of reading matter? Here's a huge collection of articles about literary forgeries. (I take no responsibility for the ad at the top of the page, which apparently changes often.) And don't forget this week's bumper crop of News of the Weird.

Yesterday I bought the best pair of headphones I've ever owned - for $19.95! It's remarkable how far headphones have come since the 1970s, largely due to the invention of new materials for making strong, small magnets.

Misconceived product of the day: Linksys WRE54G

[See June 8 to find out how I got this thing working.]

I normally have a high opinion of Linksys network products, but I've just spent a frustrating hour trying to get a Linksys WRE54G Wireless Range Expander to work, and I'm convinced the product is fundamentally misconceived.

This handy gadget should function as a repeater on a wireless network. It's just like adding another wireless access point, except that you don't have to wire it to the network hub. It connects wirelessly.

Well... There's a chicken-and-egg problem. You can't communicate with it until it's set up and working, but you can't set it up until you're communicating with it.

It should have had a cable socket in it through which you could communicate with it at all times. After all, that's how we configure other Linksys products: temporarily take the gadget off the network, cable it directly to a PC which is also off the network, and go to the specified IP address.

The Range Expander has an autoconfigure button and also a setup disk. I couldn't get either of them to do anything useful. Apparently, the Range Expander can't configure itself automatically on a network that has WEP security. (Makes sense... I don't want a stranger bringing in a repeater and putting it on my network without knowing the password.)

The setup software did mess up the attributes of my computer's wireless network connection, putting in a fixed IP address I had never seen before, and leaving me unable to use my own network until I happened to think of looking at all the TCP/IP properties.

And I couldn't configure it manually either. I don't know if I have a defective unit, defective software, or just an unforeseen situation.

So this one is going back to Fry's later this week and I'll seek another way of improving the network coverage in our house.

Too busy

I'm too busy to write anything here today, but I did revise this.

Note to the eager public: I am not "out for the summer." The University does not shut down its research labs for the summer, and thanks to research funding, I am employed 12 months this year. There have been years in which I wasn't full-time in the summer, but this isn't one of them.

Also, because there is no clear vacation policy for 12-month faculty members, if I'm not careful I never get to go away!

One brief technical note. In order to record phonograph records with my computer's sound card, I had to modify my stereo amplifier to attenuate (weaken) the signal 15 dB. (I did this with a resistor network at the inputs.) It was producing considerably more than line-level output from some records, and I've also encountered this with another amplifier (both mid-1970s vintage).

Getting music from vinyl, and even wax

Future generations will marvel at the mechanical simplicity of the 20th-century phonograph.

The first version could play and even record with no electronic components, although brass band music was the only kind that sounded very good. (Maybe that's what propelled John Philip Sousa to popularity.)

All you needed was a mechanism to rotate the disc and a needle coupled to a horn to convert the wiggles on the record into vibrations in the air.

Recording was done with the same gadget and a softer disc, so that the sound would wiggle the needle.

The 78-rpm records through the 1950s maintained interoperability with the earliest disc phonographs. Then "high fidelity" set in, with electronic recording and playback, and to this day, some audiophiles claim that nothing sounds better than a good vinyl disc. The catch is that vinyl discs don't stay good very long; they wear out. Digital audio gives us 98% (or more) of the quality 100% of the time, and that's good enough for me.

Well... Today, for the first time in decades, I can play 78s. I'm going to digitize some of my father's records from the 1940s, and even some that belonged to my great-grandmother at the turn of the last century.

Assembling the equipment for this was amusing. To play modern records we use Melody's upstairs stereo, which has a 1980s turntable and an Audio-Technica AT92E cartridge that tracks at 1 gram.

But it doesn't do 78 rpm, nor does it track old, damaged records very well. For those purposes, I've revived the Garrard 440M turntable from my undergraduate days. It originally had a Pickering V-15 cartridge with 4 grams of tracking pressure. At some point, to solve an electrical problem, I installed a P-mount adapter and a Pickering ATE-2 cartridge that was left over from another turntable.

That setup never sounded all that good because the 440M is not really smooth enough for the light tracking pressure (1 gram) of the ATE-2. Nor did I succeed in setting it light enough; I didn't have a gauge.

But...surprise...the stylus from the V-15 will plug into that same cartridge, and at the appropriate pressure (for the stylus, not the cartridge), it sounds fine. This is a surprise because the stylus assemblies don't look alike.

Pickering ATE-2 cartridge with original stylus (left) and with stylus from V-15 (right).

I had to make my own stylus pressure gauge to adjust it. This is just a seesaw made of plastic. A piece of toothpick, glued in place, is the fulcrum, and a large nut is the weight. It was calibrated with a penny (3 grams) and then the markings for the other weights were calculated (weight times distance from fulcrum equals a constant). [Note added Dec. 20: A penny doesn't weigh 3 grams any more, it weighs 2.5 grams (since 1982). I had to recalibrate the balance after discovering this.]

In use, you just set the stylus down on it and see if it balances. The plastic that I used is soft enough that it won't damage the stylus.

Having gotten that far, I ordered, from LP Gear, a 78-rpm (3-mil) stylus for a Pickering V-15, and fecklessly plugged it into the ATE-2. Audiophiles somewhere will probably scream about this, but it works fine.

So now I can play my father's 78s, digitize them with my computer, and make them sound a lot better. Click and listen to these examples, before and after. Notice how GoldWave takes out all the record noise. (Well, almost all.)

That song, by the way, is "Blue Steel Blues," composed and played by Ted Daffan, better known as the man who composed "Born to Lose." To my amazement, Ted Daffan's performance of "Blue Steel Blues" is still available on CD.

I hasten to add that I'm not a fan of country music, so this is as far as I'll go pursuing Ted Daffan. This particular record is my favorite of a stack of old 78s my father gave me, and which I played a lot as a child because I didn't have any other records. It's not a bad piece of music, and it is certainly, by now, a piece of history.

Leftover notes from yesterday

Following up some questions from yesterday's entry:

The files that I deleted did not go into the Recycle Bin because they were far too big.

Also, I did try file recovery software, but by the time the error was discovered, I had done some image processing and audio processing work with really big files, and the deleted files had been overwritten.

What's this "USB20 Device" business?

All USB 2.0 devices will supposedly run - slower of course - on a USB 1 port. But just now I plugged a new Western Digital removable disk drive into the older computer that we use for financial bookkeeping, and it wasn't recognized. Windows XP called it a "USB20 Device" and said it couldn't find a driver.

The solution? Right-click on My Computer, choose Manage, Device Manager, go down and find the USB controller, and "Update Driver" on everything that has to do with USB. Some of the drivers will get updated, and then your computer will recognize USB 2.0 devices correctly even if its USB port is the older version.

The best weather radar

I think I've discovered the best weather radar maps in the business, and as I understand it, they are free for noncommercial use. Click here to see one, and here to see the full page for the same site.

The links to these are respectively:


If you're not near Atlanta, change FFC in those addresses to the appropriate station identifier. To find your station identifier, pick the station here and notice the ID at the end of the web address to which it sends you.

Old news, but new to me: The chairman of British American Tobacco does not smoke because he considers it a health hazard. I wonder if he considers it a health hazard for his customers too.

Speaking of hazards, lead-free solder will be obligatory in electronics manufacturing in Europe after July 2006, to reduce the amount of lead that is winding up in landfills.

I've used lead-free solder, made of tin and antimony, just a bit. It's no harder to work with than the usual tin-lead mixture, but its melting point is higher (and it appears to be mechanically stronger, which is a good thing).

But manufacturers aren't going to be able to use parts made in past years unless they are lead-free. And that is a problem. For some common semiconductors, lots of people are using back-stock accumulated as long as 10 years ago.

Does this mean lots of good stuff is going to get sold off cheaply as surplus? Or just thrown away?

And will I be allowed to continue using tin-lead solder for my own experimental prototypes? I have more than a lifetime supply to use up!

Today's hot management issue: tattoos.

I want to be tolerant, but my gut feeling is to side with the Jewish moralists who contend that mutilating your body by tattooing or piercing is an insult to God's creation.

Admittedly, some modern tattoos are very elegant - such things as vines and ribbons wrapping around an arm. I doubt that those will ever get their wearers in much trouble. But a few tattoos are deliberately obnoxious, and an awful lot are just ugly and awkward-looking.

People get tattooed in order to send a message to the world, just as hippies adopted unconventional clothing and hairstyles to send a message to the world. And then they complain when the world listens to the message.

If I could invent a good tattoo-removing technology, I'd make a fortune. Too bad I'm not a dermatologist.

Today's economic issue: Buying a house can be a risky investment, especially if you go for an interest-only loan in which you are essentially betting that your house will never decrease in value even temporarily.

What made houses a good investment in the 1970s? Inflation; the value of the house wasn't growing so much as the dollar was getting smaller, so you could pay back your loan in smaller dollars.

What made houses a good investment in the past few years? Quite possibly the abundance of foolish, over-generous lending by banks. That won't go on forever.

A useful criterion is that the value of a house is what you can rent it for (because rent prices depend on people who want to use the house, not just sell it to a bigger fool) and the selling price of a house should equal about 10 to 20 years' rent, depending on condition (see here for a more technical discussion).

Now what about those people who say you should spend such-and-such percent of your income on housing, such-and-such on a car, and so forth?

Percentages like that are somewhat useful for telling you if you're way out of line. But I agree with the skepticism recently expressed by Walter Updegrave. All too easily, "you should spend X percent on housing" could be a trick to get you to spend more than you really need. Does everybody want the same lifestyle? Does everybody get the same bargains?

How I lost my files

I've figured out how I managed to accidentally delete a lot of files last week.

I was trying out a new DVD burner and thought, incorrectly, that Windows XP could burn a DVD the same way it burns a CD.

So I went into My Computer, opened up the DVD drive, and dragged the directory into it. This doesn't actually copy files onto a DVD or CD, of course; it just forewarns Windows that these are the files you want to record.

Accordingly, they show up as what looks like shortcut icons. Well, not exactly shortcut icons; the little arrow in the corner points down, not up, but you need the eyes of an eagle to see this. Anyhow, I understood them as functioning as shortcuts. They aren't really your files and folders; they are only indications that those files and folders are slated to be copied to CD. Or so I thought.

At this point I realized my DVD burner doesn't support this operation. So I wanted to bail out of the procedure and remove everything from the list of files to be burned.

How might you do that, you ask?

Right-click on them and pick "Delete"?

That's exactly what I did. And of course, when asked whether I wanted to delete, I said yes, because I thought I was just deleting shortcuts, or entires in a list of files to burn onto the CD.

No! I was deleting the original directory!

The correct procedure would have been to go to the left and choose "Delete temporary files":

I have to list this as a classic blunder in user interface design. Anywhere else, if you delete a shortcut icon, you do delete the shortcut, not the original folder. I sincerely hope this will be fixed in a future version of Windows.

The slow death of Kodachrome

Kodak is discontinuing Super 8 Kodachrome movie film. This surprised me - I thought the stuff hadn't been made for 20 years or more. Video is much better than tiny-format (sub-16mm) movie film, or so I thought, as well as being vastly cheaper and easier to work with. Well, Super 8 still has a following among fine-art filmmakers, presumably those on tight budgets. Ektachrome and black-and-white film are still made in that format, but that's not good enough; it has to be Kodachrome, which was the finest color film you could buy 50 years ago.

I think I remarked on this a few months ago: the patents on Kodachrome have expired, and if the Kodachrome process were that great, some other company would be imitating it. There were in fact two Kodachrome-like products from other companies in the 1960s, Dynachrome and Sakuracolor, neither of which lasted long. Instead, Ektachrome is the film that everybody imitates.

The best color film you can buy in 2005? In my opinion, Ektachrome Professional 100G. That's a color slide film; I don't know what the corresponding movie film is called. If you want more restrained colors, particularly for portraits, try Fuji Astia.

Meanwhile, Agfa, whose color slide film I used a lot of in the 1970s, has filed for bankruptcy. For some time they've been running a distant fourth behind Kodak, Fuji, and Ilford in the competition for serious photographers' loyalty. Agfa products have always been somewhat hard to get in the United States, and since the introduction of E-6 Ektachrome 200 (a quarter century ago!) I haven't used any Agfachrome. I like the Agfa film because it gave more separation between warm and cool colors than Kodak or Fuji films did. But Ektachrome gave me more than twice the speed with the same grain, and although it didn't have Agfa-style color, the color was generally pleasing.

By the way, the reason there's more than one kind of color film is that people don't all see colors the same way. Film renders color by using three primary-colored dyes. Unfortunately, the optimal primary colors in our eyes - the wavelengths that can be mixed to mimic all the other colors - are not exactly the same from individual to individual.

All color photography, as well as paint mixing, relies on mixing primary colors to match other (non-primary) colors. This works because of human physiology, not any characteristic of color itself; there could easily be space aliens whose vision is based on four primary colors, or ten. (There may even be a few humans with four-color systems, and there are certainly color-blind people with only two, or one.) Normal humans have three primaries, but not always exactly the same three. The difference between red and magenta may be in one place for you and another for me.

If what you are looking for is not here, please look at previous months.


Search Covington Innovations   Search Web