Short notes

Power outage: My odd experience yesterday (Jan. 30) was sitting through a half-hour power failure in the UGA Science Library. The room had big windows, and almost everyone was using a laptop, so we all just sat there and kept working. The Wi-Fi service was out for a couple of minutes and then resumed functioning, apparently on an auxiliary generator.

Snowjam: The mayor of Dalton, Georgia, points out that up there, they pre-treated their roads before the snow fell on them. Atlanta could have done the same. Two other observations: (1) The high proportion of excessively long commutes in Atlanta right now is due to the recession; people have had to take jobs a long way from home but can't sell their houses. (2) Atlanta needs a commuter railway. We can't go on putting more and more people on freeways forever.

Calendar note: A surprising amount of software (and also some human minds) in the medical and pharmacy professions seem to be stuck on the idea that a month is 30 days. If you dispense 30 pills to be taken one a day starting December 30, refillable January 30, there will be a day without a pill, because December is a 31-day month. If you assume that all 12 months of the year are each 30 days long, you'll come out ahead in February but will come out five days short for the whole year. I haven't been personally impacted by this but have come across it as a widespread practice and am wondering who, exactly, is having that much trouble doing arithmetic. I fear it's computer programmers somewhere.

Permanent link to this entry

Do you have to reboot your router a lot?

For the past year or so, we've had intermittent Wi-Fi problems: poor connections to the Internet and even computers dropping off the Windows LAN unexpectedly.

We never found the source of the trouble, but it was always cured by rebooting the router. We've had this with both Cisco and Asus Wi-Fi routers. So we got accustomed to rebooting the router every few days.

Recently I got a new perspective on the problem. A consumer-grade Wi-Fi router/access point can reportedly only handle six to ten clients, even though the instructions do not say this. The total number of clients had gotten up to 14 with our two desktops, my two laptops, Melody's two laptops, Sharon's laptop, and various iPhones, video games, Roku box, and what not.

The cure? Add another Wi-Fi access point, which in fact is a Linksys router with DHCP service turned off, connected downstream of the main router. After moving a few of the most mission-critical computers to the second access point, we've had no more problems.

We'll see if it lasts!

Permanent link to this entry

How not to handle a winter storm warning

We're snowed in today. Many major streets across North Georgia are impassable and we've been advised to stay home unless absolutely necessary. Atlanta was hit worse than Athens; not only did they get more ice, but many Atlantans have 1- or 2-hour commutes even under ideal conditions. As soon as the roads get icy, some of the cars stop moving and hold everyone else up, including the road crews that could bring salt and melt the ice. There are still people stuck on the roads who have been in their cars, and even school buses, for more than 12 hours. Some are being rescued by the National Guard.

I blame several government agencies for making bad decisions. When deciding whether to close for a winter storm, most businesses follow the lead of government offices and schools, and this time, they were misled.

A winter storm watch was issued around 5 a.m. Monday, and a winter storm warning, around 3 a.m. Tuesday.

On Tuesday morning, nonetheless, all the key people decided to have at least a partial day of school and work, even though they were under a warning. As I understand it, most of Atlanta ignored the warning completely; Athens closed schools and other facilities around midday. Then the snow and ice came, more in Atlanta than in Athens. Atlantans were still at work until 5 p.m. and then hopped into their cars for the two-hour drive home, which turned into a nightmare.

Admittedly, the storm did have a surprising aspect. It was initially forecast to go south of Atlanta. But I want to emphasize that on Tuesday morning, Atlanta didn't just have people speculating that it might snow, or TV meteorologists crying "wolf" — Atlanta was under a storm warning from the National Weather Service.

This storm was particularly tricky because the weather was warm (over 40 F) just a short time before the snow began. People were taken by surprise.

Yes, it might not have snowed after all. My friend and colleague David Stooksbury, climatologist, points out that most people don't understand the probabilistic nature of weather forecasts. No, we can't be sure it's going to snow. Yes, you should be on the lookout when we say it's likely. It's not enough to just hope that it won't snow after all. Above all, don't get so far away from home that you couldn't get back home quickly if the storm hit. That was the biggest mistake Atlanta made, collectively. The average commuting distance is so long that, effectively, everybody is out of town all the time!

By the way, people keep asking us, "Why do you Southerners shut everything down for just a little snow?" The answer is, it's not snow, it's ice. It's not powdery. When it snows in the South, the temperature hovers around freezing, or the snow hits surfaces that are above freezing, and we get a heavy coating of ice on everything. The North shuts things down for ice storms, too, — I've been up there and seen it happen.

Permanent link to this entry

Death of a newspaper

The Athens Banner-Herald, our town's more-than-a-century-old daily newspaper, has in recent years become mostly a web site, although some people still receive it on paper. The web site was advertiser-supported. That's as it should be; traditionally, newspapers have relied on advertising income for everything except "the last mile" (printing and distribution), a cost that has now gone away.

Not any more. The Banner-Herald is turning into a subscription-only web site — subscriptions (to the tune of $10 a month) required to read anything, not just premium content.

So yesterday they were publishing a list of roads known to be icy, and they wouldn't let me read it without paying money.

That did not impress me. I think that public-safety-related news items, at least, should be free to read. Indeed, there is a rival, free web site (Classic City Today) as well as another one run by radio station WGAU.

We may be very close to the time when Athens doesn't have a daily newspaper. We need to designate an alternate official outlet for legal notices and a central repository of obituaries. Then we might not need the "Banana-Herald" (as we've jokingly called it for decades) at all.

Permanent link to this entry

Snow days

Today, January 28, we had about an inch of snow, which started falling when the air temperature was about 37 F. The air quickly fell below freezing, and the result was a half inch of ice all over the streets.

This being Georgia, we don't have a huge fleet of salt trucks and snowplows, although of course there are some. So we are having a snow day. The University closed in mid-afternoon and will be closed all day tomorrow.

I have plenty of programming work to do, and no urgent need to go anywhere.

Atlanta is much harder-hit. The city is very prone to traffic jams even when conditions are ideal. Reportedly, people have been stuck in very slow-moving traffic for 8 hours or more. It's not quite as bad as Snowjam 1982 (which Melody was caught in, although I was up in Connecticut at the time), but this time we're hearing about it in real time, thanks to cell phones.

Permanent link to this entry

Subversion: Error 405: PROPFIND method not allowed

I use CollabNet Subversion (which is freeware) for version control. The error message:

Unexpected HTTP status 405 'Method not allowed' on '/xxxxxx'

Additional errors:
PROPFIND request on '/xxxxxx' failed: 405 Method Not Allowed

in fact means that the URL given for the repository is incorrect (such as leaving out '/svn'). Now you know.

Permanent link to this entry

Short notes

Work is abundant; time for writing blog entries is not. Here are a few short notes.

We have bought our first LED light bulb. It is a 120-watt-equivalent reflector floodlight bulb by Utilitech, about $32 at Lowe's, and it has a 2-year warranty. (Who ever heard of a light bulb having a warranty?) It's in a lamp in the den, pointed straight up at the ceiling, and it makes an impressive contribution to the illumination of the room.

The future, of course, is not LED bulbs but LED light fixtures. The LED lasts as long as the other electrical components, if not longer, I'm told. Right now we are in a strange transitional era, buying compact fluorescents and LEDs that are designed to fit in the sockets of old incandescent light bulbs. That era won't last long.

I've seen another wave of confused journalism about whether the recession is over. A recession is a period when GDP is falling. As soon as it stops falling, by definition the recession is over. That does not mean prosperity has come back. It only means the fall has bottomed out. Yet lots of people seem to think (and say loudly) that the recession won't be over until they're as prosperous as they were in 2005. (Or thought they were. Americans were in debt up to their ears at that time; they were spending a lot, but that's not the same as being prosperous. Debt is not wealth; it is the opposite of wealth.)

Permanent link to this entry

Yes, you can still run Windows XP

[Updated.]

People are panicking because Microsoft is going to stop supporting Windows XP in April.

Does that mean your Windows XP computer will go dead? No.

It only means that there will be no more updates and corrections to XP, and you should provide your own antivirus software, which will continue to be updated by its manufacturer (not Microsoft).

And don't expect XP to work with software and hardware newer than itself. No old operating system ever does.

Some people seem to think a Windows computer has to be connected to Microsoft over the Internet in order to work. It does not. That is not how computers work. You could continue running Windows XP even if Microsoft disappeared from the face of the earth.

As for discontinuing corrections and updates — seriously, if any major flaws still exist in Windows XP, they must not be fatal ones. (Correcting them at this point might even disrupt old software that was designed to work around them.) Microsoft has already replaced XP with Vista, then Windows 7, then Windows 8, then Windows 8.1. Do they need to continue updating a product from eleven years ago, separately from its successors?

By the way, there was one substantial flaw in XP that just got corrected. Definitely run Windows Update one last time, to get that.

In what I just said, I was talking about performance, not safety. I would be hesitant to use Windows XP on the Internet with confidential data, such as online banking or shopping, because Microsoft is no longer making security improvements and changes. It is no longer a moving target. Good commercial antivirus software may not be enough.

The most important new feature of Vista/7/8 is that you get a warning whenever software tries to make a change to your computer. Remember the Apple ads that tried to convince you that those warnings were a bad thing? They are signs of safety. If a web site, or something else, tries to tamper with your computer, don't you want to know? Switch to safety: Windows Vista or higher.

The only good reason to continue running Windows XP is to keep using old software and/or hardware that aren't compatible with newer versions. XP will run software from much older versions of Windows and even DOS that won't run under any newer version of Windows.

The change from XP to Vista broke a certain amount of old software (mostly software much older than XP itself, and/or software that never followed Microsoft's standards in the first place). The changes from Vista to 7, 8, and 8.1 were quite minor by comparison. It doesn't matter which of the three you get. And they all can run correctly written software from Windows XP, 2000, and even earlier.

If your main reason for sticking with XP is that you have a low-powered computer — maybe a single-core CPU under 1 GHz — and you don't really need to run Windows software as long as alternatives are available, consider Linux Mint, which is free, comes with lots of free software, and runs very well on low-powered PC.

Permanent link to this entry

Time bomb: batteries in telephones

For landline telephony, we mostly rely on a cordless system, but we have four traditional wired phones in different parts of the house so we can make calls during power failures (which take out the cordless system).

Today it came to my attention that two of them didn't work. One was an AT&T wall-mounted phone and one a GE, both with Caller ID.

That meant they had AA cells inside. Of course, we didn't notice that the Caller ID display went blank a while back, but we certainly did notice when one of the telephones lost the ability to dial.

The culprit? Battery corrosion.

Inquiring minds want to know a couple of things. Just why did the batteries leak? That is usually a sign of electrical abuse, not just aging. (These batteries were still within their shelf life date range.) Does the telephone try to charge the (non-rechargeable) batteries while it is powered from the line? (That would be an easy mistake to make.) Does the Caller ID draw heavy current from the batteries? Why?

Anyhow, newer AT&T phones — at least those without Caller ID — do it right. They do their memory backup with supercapacitors that are charged from the phone line and go dead if the line is disconnected more than 30 minutes. (Hardly a serious handicap, in my opinion.)

I ordered two AT&T 210's, and that's that.

But here's a deeper question. Why add so many features if all you do is add unreliability? It's hard to find a non-battery, non-Caller-ID telephone these days.

We've been through this with the "gadget cars" of the 1960s, with electric mirror adjustments, automatic headlight dimmers, etc., all implemented with flimsy 1960s electronics, almost guaranteeing that something will be non-functional at any given time. I don't need a separate set of memories on every wall phone in the house! I don't even use redial, though I can see how it could be handy.

Talking on the telephone is not my hobby. I am not looking for a full-featured telephone on which I can lavish attention as if it were a new baby. Especially not for the wall-mounted backup phones. I want the reliability and ease of use that we had 40 years ago.

So keep it simple, please.

Permanent link to this entry

How to keep a 9-volt battery from starting a fire

Don't throw a 9-volt battery into a bin or box where it might contact other batteries or metal objects unless you have covered the contacts with tape, as shown here. In at least one well-known case, an accidental short circuit across a 9-volt battery burned a house down.

I don't know how often this actually happens, but many batteries contain enough energy to start a fire.

• The hazard is greater with 9-volt batteries because they have both terminals on the same end, where they can easily touch the same piece of metal. In fact, 9-volt batteries are also hazardous in another way — it is too easy (and tempting) for a child to plug two batteries into each other, short-circuiting both.
• The hazard is greater with lithium batteries, which contain flammable material. But any battery can emit enough heat to ignite paper or similar materials.
• The hazard is greater if the battery is not dead. A completely dead battery is not a fire hazard, of course. A brand-new one is.
• Batteries removed from smoke detectors are not dead! If the detector has never sounded or reported a weak battery, then the batteries may still have 90% of their energy in them. I don't throw these away — I save them for use in less critical equipment.
• Voltage is not what starts fires. Even a 1 1/2-volt battery can deliver enough amps to heat itself and whatever piece of metal is short-circuiting it.

In fact, my own experience (and this is just my own experience; I am not making a recommendation; don't sue me) is that smoke detectors take several years to run down their batteries if the alarm never sounds. Checking for smoke takes very little energy. The reason you want a powerful, fresh battery is so that the alarm can make a loud noise for as long as possible if it needs to. That is what will demand all the energy the battery can provide.

Bottom line:

• Protect 9-volt batteries with tape (electrical tape, or other plastic tape — not duct tape, which may contain metal fibers) when you discard or recycle them, or any time they might touch metal.
• Don't throw away the batteries you take out of your smoke detector — they are still almost as good as new — use them in less critical equipment.

Permanent link to this entry

Should my telephone number be kept secret?

If you scroll through this month's Daily Notebook, you'll find a picture of a telephone with the number on it.

Two people have written to warn me that I shouldn't put my telephone number on the Internet — it's private information.

Hmmm. That same telephone number is in the telephone directory (printed on paper and delivered by AT&T) and in online telephone directories. It has even been in advertisements for my business (though I am now using a different one).

Does this cause me to get harassing calls? No. The number of harassing calls that I get from people who seem to know who I am is tiny (like one every five years or less). I can generally get the callers to reveal who they are.

We do get a lot of phone spam (robocalls, etc.) but I don't think it's from people who got my number from pictures in this blog. I think the numbers are being passed around from (shady) business to (shady) business. Lots of businesses legitimately have my phone number in their records. Ignoring a well-publicized change of area code, this number has been in my family since 1973.

Fairly recently, with the dawn of cell phones, people started dropping out of the phone book (which doesn't list cell phones) and keeping their numbers private. I confess some sympathy to this approach. My sense of etiquette tells me that, like any other demand for attention, a telephone call should not be made unless you have some prior indication that it is welcome. "Cold calling" as a sales technique leaves me cold. When done by a machine, it leaves me furious (and is generally illegal).

What do people think? Do I need to keep my telephone number secret? What do I lose by not doing so?

Permanent link to this entry

Don't battalogeîn

At church yesterday morning, our pastor, Dr. David Mills, pointed out an interesting thing. In Matthew 6, when Jesus says to avoid "vain repetition" in your prayers, He didn't exactly say "vain repetition." As reported to us in Greek by St. Matthew, he uses a rare Greek word, battalogeîn, which smacks of onomatopoeia and may be an imitation of an Aramaic phrase. It is best translated "yammer and stammer." It refers to mindless, repetitious mouthing of words or syllables.

Philological note: In a Greek dictionary, look this up as βατταλογέω or βαττολογέω. That is not an actual form of the verb; by tradition, in dictionaries, a phonological process is omitted (as it would have been in Homeric Greek) to make it easier to recognize the verb. The Greeks actually said βατταλογῶ (first person singular) and βατταλογεῖν (infinitive).

This makes it clear that Jesus did not forbid the mindful repetition of a prayer (as in liturgical worship). What He objected to was mindless noise or repetition for its own sake (as in some pagan practices of the time).

With that in mind, I'm sure I've heard battalogia that was not liturgical or prescribed at all. Some people, when they pray in public, run their mouths faster than their brains and don't make much sense, or simply pile up stock phrases without getting to the point. Even if there is no actual repetition, that is battalogia.

Permanent link to this entry

Sunspots

A fairly ordinary sunspot picture, but it shows a very large sunspot group about to rotate out of view on the right. Taken shortly before 2 p.m. EST on January 12. Celestron 5, Thousand Oaks solar filter, Canon 60Da (1/2500 second at ISO 500).

Permanent link to this entry

How "average income" can go up while most incomes go down

Further to yesterday's PBS hoopla, I have some indications that their "wages" curve is based on the average U.S. personal income (I'm not sure), and I don't think the average is the right way to summarize it. The average is too much influenced by a few people with unusually large income.

Here's an example. Imagine a village with just five wage earners, whose annual salaries are $1,000, $4,000, $5,000, $10,000, and $100,000.

Now remember your basic statistics. The mean (the average) is the sum of the values divided by the number of values. In this case, the five incomes add up to $120,000, and the average is $24,000.

The median is the item that has as many above it as below it. In this case, the median is $5,000 because there are two higher incomes and two lower ones.

The lower quartile, also known as the 25th percentile, is the item that has 25% of the other items below it. It's a good measure of what's happening low down on the income scale. In this case, the lower quartile is $4,000 because there is one below it and there are three above it.

To summarize:

Incomes: $1,000, $4,000, $5,000, $10,000, $100,000
Average: $24,000
Median: $5,000
Lower quartile: $4,000

You can see at a glance that the median is better than the average for telling you what a typical person's income is. There are lots more incomes near $5,000 than near $24,000. In fact, none of them is actually very close to $24,000 — one is much higher, and the rest are much lower.

I'll come back to possible uses for the lower quartile.

Now suppose "the rich get richer and the poor get poorer" and the distribution changes to this:

Incomes: $1,000, $3,000, $4,000, $12,000, $200,000
Average: $44,000
Median: $4,000
Lower quartile: $3,000

Look — the average went up a lot, and doesn't that mean that everyone in the village is a lot richer? Not at all. The median went down, correctly reflecting the fact that, in this example, most people ended up worse off; certainly, those in the middle of the range, and lower, did.

That is how "average income" can go up while ordinary people's income goes down.

Now look at the lower quartile. It went from $4,000 to $3,000, which is not a spectacular change, but it tells the same story as the median does. The lower quartile works a lot like the median, but it tells you what is going on in the lower half of the economic ladder. It's a good indicator of the aspiring middle class or even the "working poor," the people who are not destitute but are below the middle of the range.

There is also an upper quartile, the value that has 75% of the others below it. That could be useful for looking at the condition of the moderately rich.

The average (the mean) is a parametric statistic, meaning that it makes an assumption about how the values are distributed — that they are distributed symmetrically. The median and quartiles are nonparametric statistics, which makes them much better indicators to summarize sets of numbers that are unevenly distributed.

And in fact almost anything involving amounts of money is unevenly distributed. There tend to be a few very large values and a much larger number of smaller ones. For that reason, medians and quartiles are usually a better way of summarizing such information than averages are.

I'm expecting someone to argue that parametric statistics are "conservative" and nonparametric statistics are "liberal." No; this is just mathematics, and mathematics doesn't care whether you are a Republican or a Democrat.

Permanent link to this entry

Did PBS fail basic economics, or maybe arithmetic?

[Revised, and then further revised.]

What is wrong with this picture and caption?

First: The main "fact" being reported is obviously false. Wages may not be rising exactly in step with prices, but they certainly haven't been "largely stagnant over the past 45 years." Do you know any job that doesn't pay more dollars per hour than it did in 1969? The minimum wage was about $1 per hour back then.

Of course, if you adjust wages for inflation, they have indeed been more or less "flat." That's what we'd expect. But that's not what PBS said.

Second: The graph displays a major blunder. As the story makes clear, on the graph, the wages have been adjusted for inflation but the prices haven't.

What does "adjusted for inflation" mean? Normally, it means divided by the Consumer Price Index. So, if the inflation-adjusted "wages" line is horizontal, then it means wages are keeping up with prices, exactly as they should be.

It does not mean wages are holding constant while prices skyrocket. It means both are going up at about the same rate.

The graph is misleading the "prices" line had not been adjusted the same way as the wages. It gives the completely false impression of a great rise in prices unaccompanied by a rise in wages. In short, the graph seems designed to mislead.

If the "prices" line were adjusted for inflation, it would be perfectly horizontal and flat, since "adjusted for inflation" means adjusted for price level, and price divided by price is always 1.

Note that I have no idea whether wages are really keeping up with prices in the short term, or whether they are pulling apart. That is what we would like to know. Not for a moment would I say there is nothing wrong with our economy or the distribution of income. My point is that this graph hardly tells us anything about that. It plots two curves that are measured in totally different ways and cannot be compared.

What you want is to plot just one curve, wages adjusted by CPI, and plot it on a scale that enables us to see relatively small fluctuations in it. In fact, as plotted here, it seems to be sloping upward, which would mean that wages are rising more than cost of living — which is the opposite of what the story says. I have no idea whether any of the data can be trusted. I suspect they may have used average (mean) wage data when they should have been using the median or quartiles; that mistake would give them numbers that are unduly influenced by a few of the highest wage earners. But I don't know.

If this had been datelined April 1, I would have thought it was a joke. As it is, I can only call it a regrettable blunder.

I don't know if PBS will issue a correction. In my opinion, misreporting of facts to this degree requires a correction; we're talking about facts, not opinions. You can click on the picture above, which is a screen shot, and see whatever is now on the same web page.

I thank Mike Gay for bringing this to my attention. I claim fair use of PBS's copyrighted material for purposes of critical review.

Follow-up, Jan. 13: PBS has issued a correction in which they admit that the two lines should not have been on the same graph, so they have turned it into two graphs. This still does not result in a fair comparison because people are still going to compare the two graphs as if they meant the same thing. What they have not done is plot the two lines on the same graph either both with, or both without, inflation correction.

Permanent link to this entry

Wet

No, the rain gauge didn't quite overflow, but we did have a very rainy night and morning (today, Jan. 11), along with thunderstorms (extremely rare in January). Last week, we were having weather typical of Moscow; this week, Havana!

Permanent link to this entry

Paranoia about "personal information"

[Rewritten for conciseness.]

One of the topsy-turvy things about this crazy modern world is the increasing amount of "personal information" we're expected to keep secret. Supposedly, if the bad guys get any "personal information" about you, they'll empty out your bank accounts, steal all your possessions, and kidnap your children, and it will be your fault.

And "personal information" is more and more broadly defined. I've heard of women being cautioned not to reveal their own maiden names (or even not to use their real names at all). Most people seem to think they should keep their home address secret. And so on.

I'm old enough to remember telephone directories. I'm even old enough to remember when Social Security Numbers and even credit card numbers were not particularly secret. I understand why there are good reasons to keep them relatively private now — but let's remember, they weren't invented to serve as passwords; the assumption was that identity would be checked by other means.

Then there's biographical information. For some financial purposes, I can "prove my identity" by giving my birth date and my mother's maiden name. But are these things secret? Not at all. Much of my family knows them, of course. They're even in the Daily Notebook. And, worse than that, I'm a book author, so my birth date and parents' full names are in Who's Who in America, which you can find at any public library.

I think things are, as I said, topsy-turvy. Upside-down. Reversed. Backward.

The main problem is that the financial and computer industries would rather burden (and scare) the public than actually prevent identity fraud.

If a bank is foolish enough to give my money to a stranger who turns up and happens to know my birth date and mother's maiden name, I would conclude two things: (1) The bank has been robbed; I haven't. (2) The bank is foolish.

I also think there's been some confusion about the logic of verifying identity. If someone doesn't know my mother's maiden name, that person is definitely not me. If someone does know my mother's maiden name, at most he might be me.

That might actually be enough to deter thieves working from lists of names and addresses or collections of stolen credit cards. My mother's maiden name is not in my wallet or in any database of credit card transactions that a thief might crack into. They're not going to spend 15 minutes searching my web pages if they can just move on to the next name in their stolen list.

The birth date and mother's maiden name also distinguish people with similar names. Anyone giving my birth date and my mother's maiden name is definitely claiming to be me, not the other Michael Covington who went to Georgia Tech and whom I have narrowly missed meeting several times.

I'm not saying you shouldn't be careful. Here are some suggestions:

(1) Without being unduly secretive, you don't have to make things easy for an identity fraudster. I don't call a lot of attention to my home address or various things that might be the answers to security questions, even though I'm not secretive when they need to be mentioned. I am more careful about what I put into databases (which might be breached on a large scale) than about what I say in my blog or on Facebook.

(2) When you have to choose security questions for a computer account, choose relatively unusual ones (not "mother's maiden name"). But not questions that don't have clear answers. I wonder how many people can't specify a definite "name of first girlfriend." Fortunately, the candidates for that title are not reading your answer.

(3) Consider deliberately misspelling or falsifying the answer to the security question. Do this with great caution, because, remember, the purpose of the security question is to rescue you if you forget your password. Don't just give yourself another password that you will just as surely forget!

(4) If you are the victim of identity fraud, remember that you are the victim, not the criminal. Remind the bank, or other business, that it chose to take risks.

(5) Don't be paranoid about rare crimes. Somehow, everybody's afraid of their children being kidnapped, but it's almost like a superstitious fear of witches. Apart from divorce custody disputes, kidnapping is rare. Burglary is common, so don't do anything to help burglars (such as pre-announcing your vacation on Facebook). But don't imagine that one geotagged picture is going to get your children kidnapped. They are much more likely to be run over by a drunk driver (and so are you).

Bottom line: We are under too much pressure to keep basic personal and family information secret, and I'm resisting the pressure. We need better technology, not paranoia. And it's the bank's job to keep itself from being robbed.

Permanent link to this entry

Homedics massage chair:
"Power" and "Demo" LEDs never light

Today's electronic repair has a touch of the outré.

I was repairing a broken power cord in a shiatsu massage chair that belongs to the gym where Melody and I exercise, and I decided to check it thoroughly.

The "power" and "demo" LEDs had never, to my recollection, ever lit up, even when turning on power and running the demo.

I had a look inside, and to make a long story short, those two blue LEDs were installed backward, with reverse polarity. I turned them around (disobeying the orientation stenciled on the circuit board), and now they work.

Somewhat misleadingly, though, they light up all the time, not just when power is on or the demo is running. This seems to be the way it was designed, though not the way it was built!

I added an explanatory label on the back of the handheld control unit.

The moral: If a problem has been present since a piece of equipment left the factory, suspect the factory!

Permanent link to this entry

Are mail-order eyeglasses any good? One experience

I've worn glasses almost all my life and, as age has reduced the focusing ability of my eyes, I now need different glasses for different purposes (driving, astronomy, computer work, etc.).

(An aside: Everyone loses focusing range as they get older; it doesn't mean I'm going blind. It just means that my eyes can't change their focal distance the way they used to. When I was young, my eyes autofocused and my camera didn't; now it's the other way around.)

Anyhow, I normally get my glasses from a fine local optician to whom I recently switched when the previous one (also excellent) went out of business due to a retirement.

But, just to see what would happen, I decided also to order a spare pair of computer glasses from Zenni Optical in China. Zenni's prices are really low. If you don't want anti-reflection coatings, you can get a pair of glasses for well under $10. I had mine coated and still was out less than $25, including shipping.

So... was I satisfied?

Well...

Optically, the cheap glasses are just fine. They were single-vision lenses for near vision. I would be more worried if the lenses had been bifocals or, even more so, progressives, which require real expertise to fit.

The fit of the frames was just awful and Zenni's online instructions for adjusting them were not adequate. I eventually had to call upon the local optician for the touch of the master's hand. Please do not expect opticians to adjust glasses that you bought somewhere else free of charge. I would have been glad to pay for the service.

Update: The real problem was that the frames were much too small for my face. Most of Zenni's frames are designed for small faces. A later pair of glasses, using their largest aviator-style frames, was excellent and fit right out of the box. Check the size of their frames and measure the pair you are using now, for comparison.

Bottom line: Within limits, you get what you pay for. I will be comfortable using Zenni for spare glasses (at $21 a whack, delivered!) but not for the glasses that I most rely on. For those, I want the expertise of a local dispensing optician and a major manufacturer (I prefer Zeiss).

Permanent link to this entry

Cars (and other things) in the cold

[Updated. I thank Paul Franklin for more suggestions!]

Georgia is about to have a remarkable cold snap, with temperatures a long way below freezing for several days. Here are some hints that my Facebook friends (and others) can pass along.

• Car windshield washers are damaged by freezing and can be expensive to repair. Most of the windshield washer fluid sold in Georgia is, for some strange reason, only rated for use down to 32 F. We have weather colder than that every year, but the 0-degree fluid is hard to find. I found some at O'Reilly Auto Parts. An alternative is to fortify the existing fluid by adding isopropyl alcohol (rubbing alcohol), which I've done many times in the past. After putting in the low-temperature stuff, be sure to use the windshield washer briefly so that it will circulate into the pump.
• Car batteries often succumb to cold. This is a good time to get your battery load-tested and see if it's nearing the end of its life. Also, if you have a car that hasn't been driven in several days, crank it up and drive around (at highway speeds, not in stop-and-go traffic) for at least 15 minutes to charge up the battery. This may make the difference as to whether it will start on Tuesday.
• Water pipes can freeze and burst, even in places where this has never happened before. Cover your outside faucets. If you have a basement or other incompletely heated area with water supplied to it, leave the faucet dripping so that enough water will circulate to (hopefully) prevent freezing. Open the cabinet doors under the sink so that room air can get to the pipes.
• Keep your furnace set to a reasonable temperature rather than waiting until the house gets really cold, and make sure you are actually giving enough heat to the whole house, even if this requires setting the thermostat relatively high. For example, if you are using a fireplace in the room with the thermostat, you'll need to set it higher in order to heat the rooms that don't have a fireplace adequately. My thermostat is over a laser printer and the same principle applies.
• Electric lights and appliances give off heat, and this can be useful. (Just leave some lights on, and you'll have more heat. Also, this could be a great time to run a self-cleaning cycle on the oven.) But don't use appliances in abnormal ways, such as stove burners with nothing on them — you risk starting a fire.
• Carbon monoxide kills! Don't leave a car engine running or use a gasoline-powered generator near the house. (Even outdoors, a few feet from the house is not necessarily enough.) And follow instructions for kerosene heaters — they are not for indoor use unless there is ample ventilation.

Permanent link to this entry

A thin moon in the twilight sky

I snapped this picture quickly, in twilight, on January 3, using my Olympus 600-mm f/6.5 lens at f/8 and my Canon 60Da on a fixed tripod. The air was cold and windy and I had just brought the lens out from the warm house, so atmospheric steadiness was against me.

Permanent link to this entry

A point about opinions

On any controversial issue, it is not news to me that you have an opinion. I knew that you had an opinion already, and I'm not taking an opinion poll. There are times when an opinion poll is needed, but it's not the purpose of rational discussion.

What matters in a discussion is whether you can tell me (or the group) anything I (or they) didn't already know.

I've just had a frustrating time on Facebook. I posted a news story about a gun accident and asked people not to debate gun control. A couple of them did anyway, along with complaining about "not being allowed to speak my mind," and I deleted the whole message thread, then posted an explanation. The same thing happened again in response to the explanation. I deleted that thread too and posted a further explanation.

The real problem with the gun control debate is that each side knows almost nothing about the thoughts and concerns of the other side. They think they do, but they don't. They have some learning to do — and learning is best done with the eyes and ears, not the mouth.

Permanent link to this entry

Short notes

A thought about debates: Forceful arguments and combat do not make your opinion true. Weak arguments and timidity do not make your opinion false. I'm looking for truth, not victory.

A related point: I'm not required to have an opinion about everything I ever hear about, just because it's in some newspaper or someone has mentioned it to me. Not everything affects me, and in many cases I'm never going to hear enough about an event to have an informed opinion. We should beware of attempts to stir up our emotions by giving us half-formed accounts of obscure incidents that seem unfair. That's where prejudices come from.

More farewells: I sadly note the passing of Jon Liston, who was a manager with UGA Physical Plant and did an outstanding job of taking care of the building I worked in, and of Dick Mendenhall, who interviewed me on the radio back in 1997. (Click through and you can hear it.)

Household hint of the day: "Grip" shelf liner (the kind that is like a coarse rubbery fabric) is useful far beyond lining shelves. Put a small piece of it under anything you don't want to slide across the table (particularly useful with small pieces of electronic equipment). Or glue small pieces to the underside of the equipment, or to its rubber feet, if it has them but they aren't performing their anti-skid function. This is an easy way to make small gadgets feel a lot more solid and substantial.

This stuff is available at all large stores; Target has more than the usual number of colors and sizes; see also the Amazon link herewith.

Permanent link to this entry

No such thing as a plain light bulb

Today I'm replacing light bulbs. Look at how much difference in height there is between these two CFL bulbs. The smaller one is newer and is admittedly a lower wattage, though almost the same brightness. The difference in size means that one of them sticks out of a track lighting fixture, and the other, thankfully, does not.

Lesson? I have to shop for light bulbs by brand and model number, not just type and wattage. Sizes, brightnesses, and colors are no longer standardized, the way they were in the Good Old Days.

Or maybe Not-So-Good Old Days. Traditional incandescent bulbs are big energy wasters. They convert much more electricity into heat than into light.

In fact, the incandescents that you can buy today (and which cost $3.50 instead of 25 cents) are actually high-efficiency halogen incandescents mounted in ordinary-looking bulbs. That's right — there's a bulb-within-a-bulb, as you can see if you look at a clear one. They're also likely to be marked, "53 watts, 60-watt equivalent," meaning they get the same amount of light from less electricity.

Incandescents are very vulnerable to variations in voltage. Voltage is the electrical equivalent of water pressure, and my house is generously provided with both. The voltage here is usually 126 V, and a 120-volt bulb burns brighter than normal but also has an appreciably shortened life. Back in the day, I used to buy 130-volt bulbs. CFLs, fluorescents, and LEDs regulate their voltage and aren't affected (much) by the higher voltage here. Voltages above 125 (supposedly the maximum) have been common at least since the 1980s. Before you ask — no, there isn't anything simple you can install in your house to lower the voltage slightly, the way our water pressure regulator lowers the water pressure.

Permanent link to this entry

Return to twentieth-century elegance

My first electronic project of the New Year was to refurbish and put back into use a telephone that we acquired as newlyweds in 1982.

We already had one of our modern cordless phones next to the bed in the bedroom, but I found it unduly hard to answer in the dark — the buttons don't light up, and there's no way to see which button to press. Given time, I could learn how to find the "talk" button in the dark, but that's just the sort of thing I might forget in an emergency. Also, continuing to think of emergencies, the cordless system requires AC power (at the base unit) and does not work during a power failure.

So we returned to 20th-century elegance. Telephones were more ergonomic then. This one can be answered in total darkness — no buttons to press — you just pick up the handset and there you are! (The handset is slightly asymmetrical so you can feel instantly if you've grabbed it the wrong way, even before your ear tells you.) The telephone doesn't light up at all, but I could dial in darkness, if I had to, since the buttons are in the standard arrangement.

This Western Electric 2500 telephone, made in about 1976, was a tiny bit old-fashioned even in 1982; most people were getting Trimline phones, which are more self-contained and do light up. We got a Trimline initially but had to return it because it wouldn't work with an acoustical-coupler modem that we used to connect to my employer's computer at a whopping 0.0003 Mbps (300 baud).

In those days, telephones were owned by, and rented from, the telephone company, in our case Pacific Bell (in suburban Los Angeles). Very soon afterward, though — 1983, I think — the FCC required telephone companies to change their rules so that customers could use privately owned telephones. Pacific Bell sold us our rented telephones (and issued us stickers to attach to them, saying so) and we got to keep them.

Getting it back into usable condition consisted mainly of cleaning contacts with Caig fluid. I also verified that it will dial all ten digits correctly (old telephones sometimes don't). After calling my cellphone, which exercised about half the digits, I finished the test by calling Kroger Pharmacy and putting in 0123456 and 9876543 as prescription numbers, listening to the computer reading them back, and then cancelling each one. I did not have an easy way to test * or #.

For more authenticity I made a number card with a Selectric typewriter, on card stock, and then cut out a clear plastic window to cover it. This isn't perfect; I used 1970s technology, but telephone companies actually printed the number cards with modified adding machines. At least that's what it looked like they were doing the last time I saw it done, in New Haven in 1980 or so.

By the way, the expression "hang up the telephone" actually goes back to an earlier generation of technology, the candlestick phone, on which you actually "hang up" the receiver on the side of the candlestick. (So do the terms "on hook" and "off hook," as well as the more technical term "switchhook" denoting the big multicontact switch in every telephone that turns it on or off.) A few candlestick phones were still in use in my childhood, but I can't recall ever seeing one in use (not counting the replicas that were popular in the 1980s). Indeed, if you have a 1930s candlestick phone in good condition, it will still work today. Analog phone lines haven't changed since the 1920s, except by adding more capabilities.

Permanent link to this entry