Michael A. Covington    Michael A. Covington
Books by Michael Covington

Previous months
About this notebook
Search site or Web

Daily Notebook

Popular topics on this page:
Linksys WPSM54G or PSUS4: Windows 7, Vista 64, XP, Linux
Review of xFountainPens.com X750 "Legislatur"
How I teach time management
Safe Windows 7 setup recommendations
Moon (Tycho)
Moon (Janssen)
Many more...

For more topics, scroll down, press Ctrl-F to search the page, or check previous months.

If you'd like to show your appreciation of the Daily Notebook, please consider making a contribution to my department at the University of Georgia. Thanks!


Two financial tidbits

Shopping for health or dental insurance? Look at ehealthinsurance.com. Insuring a young person for a few months between jobs or schools is surprisingly inexpensive.

Cathy explained to me why option ARMs came into existence. An option ARM, you'll recall, is a mortgage where you have a choice of making a regular payment, or an interest-only payment, or an even smaller payment that doesn't even cover the interest. People got burned trying to buy houses this way that they couldn't afford.

Well, the legitimate use of option ARMs is that many of the richest people — movie stars, Wall Street wizards, and so forth — have very variable incomes. They want to pay half a million dollars once a year and just a few hundred dollars per month the rest of the time. For them, an option ARM is a winning strategy. The tiny minimum payments serve merely to reassure the lender that the customer is still there.

So if you get a half-million-dollar bonus once a year, go ahead and get an option ARM. Or, better yet, pay cash for the house.

That's all for September. October is right around the corner!


North of the border

If this does not fill you with patriotic pride, you are not Canadian. Which may be just as well...

Si ceci ne vous inspire pas avec patriotisme, vous n'êtes pas canadien. Qui n'est pas problème...

More details here.

Plus de détails ici.


Web spruce-up

Tonight we're working on web pages other than this one. The "new look" (barely different from the older one, but more consistent) will début first on my consulting page and the 404 (error) page. Then, when we're sure it's right, we'll propagate it to a lot of other pages. Main features include a centered fixed-width layout (so pages don't sprawl out on wide screens) and improvements to the navigation buttons. When we get the appearance right, we're going to go back and revise a lot of content.

What you see right now is not final. If you see problems, don't worry — we'll be fixing things all week.


The standard error of a proportion

I think the concept of standard error of a proportion should be given a lot more emphasis in statistics courses and handbooks. I had a use for it in my research the other day.

Recall that when you estimate a quantity from a sample, you get a mean (average), a standard deviation, and a standard error. The standard error, computed as

standard error of mean = standard deviation of sample / sqrt(N)

where N is the number of items in the sample. It tells you how much you should trust the mean. Obviously, a larger sample gives you a better estimate of the mean of the entire population. A smaller sample is more likely to be inaccurate. The standard error is the expected standard deviation of all estimates of the mean from all samples of that size from that population.

Assuming a bell-curve distribution of possible sample means — which is a legitimate assumption regardless of the distribution of the values being sampled — it follows that the 95% confidence interval is from 2 standard errors low to 2 standard errors high. If you get a sample mean of, say, 3.4 and a standard error of the mean of 0.1, then you are 95% sure the true population mean is between 3.2 and 3.6.

That much is in every statistics book. But what if, instead of measuring a quantity, you're measuring a proportion? That is, what proportion of your sample has a particular property (a yes-no property such as being male, not a quantity such as height)?

The property can be thought of as a variable whose values are 0 and 1. Its standard deviation turns out to be, necessarily, sqrt(p(1—p)) where p is the proportion. So:

standard error of proportion = sqrt(p(1—p)) / sqrt(N) = sqrt(p(1—p)/N)

And the 95% confidence interval ranges from 2 standard errors low to 2 standard errors high. For example, if you sample 10 University of Georgia students chosen at random and 6 of them are female, then p = 0.6, N = 10, the standard error is 0.154, and the 95% confidence interval is plus or minus 0.308, which means that in order to draw a conclusion, you need a substantially larger sample.

There are many web pages about this. Search for "standard error of a proportion."


Re-translating the Catholic Mass

The Roman Catholics are making a new English translation of their missal for use in the United States. This interests me both as a Christian and as a linguist.

Although I'm a Baptist (or, perhaps more precisely, an evangelical who belongs to a Baptist church), I've learned a lot about worship from the Catholics. Baptist worship is always informal; Catholic worship is liturgical and has changed relatively little for more than 1500 years.

Until about 1965, Catholic services were in Latin. The first one I attended was in the fall of 1968 and was in English, but using the first official English translation of the "old" liturgy, which you can read here. (Although billed as "revised," it is very little changed from what they had used for the past several hundred years.)

Many words and phrases stuck in my mind. Among them:

The Lord be with you. ... And with your spirit. (A very ancient greeting, from, if I recall correctly, the third century.)

Lord, I am not worthy that you should come under my roof; speak but the word and my soul will be healed. (Matthew 8:8)

Lamb of God, who take away the sins of the world, have mercy on us. (See John 1:23. The grammar of this one puzzled me because it is relative clause in the second person. The Anglicans translated it "who takest away.")

Big changes came in 1970. The liturgy of the Mass was revised to have a choice of four or more eucharistic prayers, instead of just one, and to trim down the prayers at the beginning, among other things. Some say it showed Protestant influence — it seems to have toned down the distinctively Catholic doctrine that the Mass is an actual sacrifice rather than just a commemoration of Jesus' past sacrifice. Being a Protestant, I don't object to this, except for some concern that the Catholics may not have known what they were getting.

More importantly, even the unaltered parts were re-translated into simpler English, which you can see (in part) here (left column). The passages that I quoted became:

The Lord be with you. ... And also with you.

Lord, I am not worthy to receive you, but only say the word and I shall be healed.

Lamb of God, you take away the sins of the world; have mercy on us.

The distinctive style seems to have been lost, and the Bible references are less evident.

Now there's a new translation about to be adopted (here, right column), and the distinctive style is back, though it's not a complete reversion to 1965. Here's what they have now:

The Lord be with you. ... And with your spirit.

Lord, I am not worthy that you should enter under my roof, but only say the word, and my soul shall be healed.

Lamb of God ... you take away the sins of the world; have mercy on us.

Well, two out of three isn't bad. In the third one, the translators must have felt that second-person relative clauses aren't actually grammatical in English (and I'm not sure I've ever seen one anywhere but here). The reason the translation is not a complete reversion to that of 1965 is twofold: parts of the Latin text were changed in 1970, and even the unchanged parts weren't translated as elegantly in 1965 as they might have been.

Interestingly, another of my favorite passages:

Let us proclaim the mystery of faith: Christ has died, Christ is risen, Christ will come again.

isn't in the Latin at all; it's an addition approved only in the United States.

Batten down the hatches...

I had a good day of programming yesterday, but today (Saturday) is a UGA home football game, and that means the University is nearly shut down for 48 hours. I hope that this time there won't be an attempt to vandalize my lab, as there was two weeks ago.



The University is in the late-September frenzy and I spent four solid hours answering urgent e-mail, which came in as fast as I could respond to it.

Part of the cause of the frenzy is that we have another home football game this weekend, and we essentially have to shut down for 48 hours, starting before the end of Friday. But all the work still has to be done.

So I have nothing for you to read today. Instead, read this essay about Keynesian economics. Informative, but not to be swallowed whole.


It's Wednesday! Do everything today!

Actually, this entry has Thursday's date on it, but was written on Wednesday. I'm amazed at how activity at UGA, and the population of the campus, and my incoming e-mail from everywhere, all peaks on Wednesday. There must be people out there with a one-day work week...

Welcome home, Melody! Melody stayed in Marietta an extra day because of flooded roads.

Jupiter again

Pete Albrecht worked on my September 7 Jupiter image to correct some atmospheric dispersion and reduce grain. Here's his handiwork:

Meanwhile, on the evening of the 22nd, I got another one, not quite as good, but presentable:

We had relatively steady air because the flood-drenched ground wasn't as warm as usual.


Safe Windows 7 setup recommendations

The University of Georgia has received its volume-licensed copies of Windows 7, and we're starting to deploy it. I am hearing very good things about its performance — apparently it has the reliabilty of Vista with speed greater than Windows XP.

Only one disappointment: a Windows 7 Enterprise setup disk (which we got) can't upgrade Windows Vista Ultimate. We'd have to wipe and reinstall the whole computer, and right now I can't spare the time.

IAI lab technician Ananta Palani has put together the following advice on how to set up Windows 7 securely:

(1) Enable 'Always Notify' in User Account Control (UAC): Control Panel > Action Center > Change User Account Control settings (on left sidebar) > Always notify (slider all the way at the top). (http://www.istartedsomething.com/20090613/windows-7-uac-code-injection-vulnerability-video-demonstration-source-code-released/)

(2) Do not enable the "secret" built-in Administrator account; it's not protected by UAC and is therefore too dangerous (it takes you all the way back to pre-Vista days).

(3) Use a 'Standard' user account after initial setup is complete. (http://unixwiz.net/techtips/win7-limited-user.html)

(4) Disable 'Teredo' protocol. (http://technet.microsoft.com/en-us/library/cc722030%28WS.10%29.aspx)

Cloven in twain

As this is written (evening of the 22nd), Melody is OK, but still trapped in northwest Atlanta. North Georgia is cut in two by the Chattahoochee River, and there are, as yet, no reliable bridges across it. (Most of the time I-285 crosses the river; today the river crosses I-285.) Melody is having a good visit with her aunt, but Cathy and I miss her.


Short notes

Atlanta is in fact having the worst flood in more than 100 years, and there have been fatalities. Because Atlanta is hilly, there are scattered places throughout the city where water is 9 feet deep, and roads are blocked. But Melody came through OK.

Software product of the month (next month): Microsoft's free antivirus program. I hear it's much less of a CPU hog than other antivirus programs. I'm torn between three thoughts: (1) Microsoft ought to know how to do a good job, because they know the OS better than anything else; (2) antivirus ought to come with the OS anyhow; but (3) there always ought to be more than one antivirus software maker, preferably at least half a dozen, so that a virus can't succeed by getting past just one of them.

Finally, either there's been a mistake, or this is the most expensive book on Amazon. The Amazon customer reviews, all facetious, are funny. (The price seems to be real; here it is in euros, on the publisher's site, in German.) Thanks to Pete Albrecht for the link.



A wet kind of drought

For several years, all we've heard about, locally, is The Drought. Until recently there were heavy restrictions on outdoor water use in Athens. We got some respite last winter, then had a dry spring and summer.

But the past week has been the wettest I can ever recall, with more than an inch of rain per day for several days. And the end is not yet in sight. In two more days we will have completely made up our "deficit" for the year (as if rain were something you borrow from investors).

Atlanta has been hit harder than Athens. Melody is over there right now, attending a Photoshop seminar, and may have to stay an extra day to wait until the roads are clear. "Spaghetti Junction" was closed this morning, as was part of Highway 316 (the main road to Athens), and there was actually a fatality when a car was swept away by water on Highway 29 near Pleasant Hill Road (not on our usual route, but close).

In Athens, the main things I notice are: (1) younger drivers don't know how to drive on wet roads; (2) buildings have been built in places that may be more flood-prone than the builders realized. No real destruction here (yet), but plenty of fender-benders.

If, at the end of this, they're still saying, "conserve water because of the drought," I'll be convinced of what I've long thought: It's not so much a drought as a lack of properly built facilities.


A bit of economic judo

Even though I'm a textbook author, I don't like to burden students with high-priced textbooks. Accordingly, as supplemental reading for my newly re-engineered natural language processing course I've specified the "obsolete" 7th edition of Fromkin and Rodman (about $9, used, and easy to find) instead of the 8th edition (about $90, new).

More college professors should consider this method of cost control. I'm sure the new edition is somewhat better than the old one, but not 10 times better. Linguistics is a rapidly advancing field, but not so much that an introductory course is noticeably different after just five years. I don't have to make people buy the new edition just because it exists.

Gold is worthless

I don't agree with financial pundit Dave Ramsey about everything, but he made a really good observation on this radio show this week: Gold is worthless. Using gold as money is just as arbitrary as using government documents as money. Gold it not very valuable to human beings — you can't eat it or heat your house with it. It's just that, for a long time, people chose to use gold as a medium of exchange. It could just as easily have been wood, or rocks, or whatever.

(A correspondent points out that gold has one advantage over government documents — the government cannot increase or decrease the supply of it.)

Ramsey has no sympathy for alarmists who say that the economy is going to collapse and people will go back to using gold. When economies actually do collapse — such as after Hurricane Katrina — people trade necessities, such as water and fuel. Gold is even less valuable then than at other times.

Except for occasional exceptional runs (one of which is just ending), gold is not a good investment. Now is a great time to sell gold, but a terrible time to buy it.

On the Google Books settlement

The purpose of books is to be read. I support Google's fast-growing archive of digitized books, and especially their efforts to scan and preserve old out-of-copyright books, as well as all the copyrighted books they can license. They even have a scheme to offer paper copies printed and bound on demand in big bookstores.

The problem, of course, is what to do when a book is still in copyright but the owner can't be found.

The Department of Justice sensibly comments that Google's proposed settlement doesn't solve the problem. In their words:

The central difficulty that the Proposed Settlement seeks to overcome — the inaccessibility of many works due to the lack of clarity about copyright ownership and copyright status — is a matter of public, not merely private, concern. A global disposition of the rights to millions of copyrighted works is typically the kind of policy change implemented through legislation, not through a private judicial settlement.

That is: Change the copyright laws to solve the problem.

I support Jeff Duntemann's proposal that all transfers of copyright should be legally recorded, like real estate sales, so that anybody can do a title search at any time.

I also think the duration of copyright should be a lot shorter. I'm old-fashioned enough to believe, as legal theorists do, that copyright was created by the government, and that it is distinct from the author's "moral right" to be identified and given credit. Copyright is a temporary monopoly on publishing, intended to encourage publication by making it profitable. It is not intended to render old books un-reprintable.

In recent decades Congress has kept extending the duration of copyright basically so that Walt Disney's early work will never go out of copyright. (I'm not kidding.) That's not how copyright law is supposed to work. I think 20 years is the right duration, because for 20 years after a work is created, you can still find people who were involved in its creation, and negotiate with them. Thanks to Disney's lobbying, some copyrights can now last as much as 120 years, and after that much time, there's no hope of doing any business with the creator of the work or his immediate representatives.

A compromise might be to require frequent, relatively expensive renewals if an owner wants to keep an old work copyrighted. Let the copyright expire quickly (like within 10 years) if the owner vanishes.

The new face of credit cards

New credit card regulations are about to take effect, and card issuers are starting to change their offerings.

The biggest change, of course, is that it's going to be somewhat harder to get a credit card. In fact, one issuer, PartnersFirst, is reportedly not using FICO scores — instead, they're going to evaluate each application individually, just like in the old days. (It has been established, by the way, that customers aren't willing to pay annual fees — that particular nuisance isn't going to come back.)

Second, lenders no longer have an incentive to give you different interest rates on different parts of your balance (such as higher interest on cash advances), because they can no longer direct your payments entirely to the low-interest portion. Thus the new Bank of America offering has the same interest rate on all transactions, and I think that's going to be normal. (Reported here, but without a permalink.)

Third, interest rates are going to be tied to the prime rate, because that's the only way they can vary; card issuers can no longer change a rate when they feel like it. And there's the rub. Prime plus what? In my opinion, prime plus 5% is a bargain (which I haven't seen); prime plus 10% is acceptable; prime plus 15% is high; and prime plus 22% (the highest I've seen) is extremely high. Remember that the prime rate is 3.25% right now but is normally more like 8%. Whether it will ever again go as high as it did in 1980 is uncertain; those were strange times, and the prime rate hit 20%. (Historical data here.)

That, incidentally, is when adjustable-rate mortgages (ARMs) were invented; people expected (quite reasonably) to refinance in the future at a lower rate. Taking an ARM when rates are exceptionally low, as people were doing three years ago, is folly. In fact, I think a lot of what's wrong with the economy right now is that people did things backward — borrowing heavily during prosperous times and scrambling to get out of debt during the recession.


Chaos on campus

As I write this (Saturday, Sept. 19, 11 a.m.), the University of Georgia is going to be without water for several hours because a water main has broken in the middle of the campus. Fortunately I'm not there. I hope someone gets some good pictures of what must be Athens' fastest-growing lake...

A week ago, there was a home football game, and so-called football fans (apparently our own, not the visiting team's fans) seriously damaged the campus, leaving behind tons of trash as well as assorted destruction. (One picture here.) One of the many vandalism incidents appeared to be targeted specifically at my lab, although fortunately it didn't do as much damage as it could have; I can't discuss the details yet.

Why do people who "support" our football team behave as if they were trying to destroy the University?


Three laws of troubleshooting

Today I gave an informal talk to my electronics students on "what to do if the robot doesn't work." It boils down to three principles:

(1) Paraphrasing Peter Pan, ya gotta disbelieve. That is, you have to be able to doubt your assumptions, question them, and look for likely alternatives. After all, the machine always works the way it is supposed to; it doesn't violate the laws of physics. If it doesn't work the way you expect it to, something is wrong with your beliefs about it.

(2) Go the extra inch. Just a little extra work — taking careful notes, double-checking the wiring, and so forth — can save you a gigantic amount of time. I don't ask you to go the extra mile, just the extra inch.

(3) Localize the problem. Instead of "the robot doesn't work," turn the problem into "the left servo motor doesn't work" or something like that. Think of the machine as a connected set of causes and effects. Try to isolate the part that doesn't work.



What's wrong with a lot of web sites

Melody and I are sprucing up CovingtonInnovations.com and have been thinking about web design.

We think that some web designers have lost track of what the Web is, and are trying to make it either too much like TV, or too much like application software.

There is a place for web pages that resemble TV, and there is certainly a place for software on the Web. But let's not forget that the Web is mainly hypertext — that is its strength and its normal use. People are looking for written messages with links that work in the normal way. That's also what search engines are looking for.

Examples of pages that are far too much like TV include Behr Paint and BestBeds.com. Not only do they bore the viewer (who expects to control the pace rather than wait for a video to play), but they also conceal most of their contents from Google.

Pages that are too much like application software are common. I don't mind software on the Web — it's a great thing — but as far as possible, web sites, especially complicated ones, shouldn't violate the rules of web browsing. They shouldn't refuse to go back if you press the back button, or lose all your data if you try.


What if they never send me a bill?

A few days ago, I got a somewhat obscure e-mail message saying that one of my credit cards from a particular bank had gone paperless, i.e., I would no longer be receiving paper statements.

(It was a card that I never use. Of course I never get a statement in months when I don't use it. So in fact I have never received a statement from this one.)

Anyhow, it took me a while to figure out which card this was in reference to, and I eventually logged in to the right place and turned paper billing back on.

But what if I had used the card, or someone had put a charge on it by error or fraud, and I'd never heard about it? Can they penalize me for not paying a bill that they never sent me?

E-mail is a lot less reliable than postal mail. A lot of e-mail never reaches its destination because of spam filtering or technical problems.

So I'm not pleased when a credit card issuer announces, all on its own, that if I owe them money, I'll never know it unless I log on to their web site.


Very short note

I'll write more tomorrow. Wednesday (when this is written) really is "hump day" — I wonder if some people somewhere are working a one-day week. Why else is the University so much more crowded on one day than all the others?

By the way, if you look at economic indicators, it's as if someone has flipped a switch. There are now analysts who expect the recovery to be as sudden as the crash.


Short notes

Economy: Mr. Bernanke says the recession is over, there are lots of good economic indicators — so when will the University of Georgia return to full funding? Two or three years from now, maybe? We depend on the previous year's tax revenues, so we're always behind. There are only two conditions under which state universities get their budgets cut: when the economy is weak (which reduces state revenue), and when the economy is strong (and people feel less of a need for education).

Based especially on what I hear in popular financial advice columns and the like, I sense that Americans are continuing to turn against higher education. Admittedly, we had too many people borrowing too much money to attend too many years of college, but let's not throw out the baby with the bath water. We're not going to go back to being a nation of factory workers who don't need to know much. That era is over.

Investing: If anybody followed in my footsteps and invested in FRESX, a real-estate mutual fund, here's a handy tip. FRESX is only priced at the end of each day, but it behaves very much like RWR, which is exchange-traded. You can look at RWR to estimate what's happening to FRESX in the middle of the day.

When does the Space Shuttle look like a comet? When it releases "waste water" (i.e., urine). Click through and see the picture.


Short notes

Sometimes I do something right: Everybody's saying the stock market rally won't last, but now is the time to invest in real estate, so I pulled a small investment out of SPY (an S&P-500 ETF) and put it into FRESX (a real-estate mutual fund). Presto! FRESX went up 3.4% in one day.

Timely idea (thanks to Sharon for the link): Keeping women safer by educating men. There is a whole "men's entertainment" culture that encourages sexual predation. Down with it!

Yesterday's "extra" entry about time management is a hit. Read it now if you haven't already.



How I teach time management

[Updated and corrected. Garish color added at Jeff Duntemann's suggestion...]

I supervise lots of graduate students who are doing either of two things: working as a 1/3-time research assistant in addition to their studies, or writing a thesis. Often, their time-management discipline is not initially up to the task. Maybe some readers of this blog will find it useful to know how I deal with that.

I don't actually teach time-management techniques (scheduling, calendars, etc.). Students already know how to keep a calendar. Beyond that, many of the techniques of the business world seem to assume that every hour is interchangeable with every other hour. That is true only if the work is nowhere near your intellectual limits. When real creativity is required, one hour is not interchangeable with another. You have to know when you're "on a roll" and when you're not. But "inspiration" doesn't come until you've started working, so don't wait for it. Just acknowledge that some work sessions are going to be more productive than others. Preferably, you should be able to choose between different aspects of your work so you can do the part that most fits your inclinations at the moment.

What I have to instill is the habit of working on your own. And right at the start, most students arrive miseducated by our educational system. They've been trained to think that work shouldn't be done until the night before it's due, and that doing 70% is enough.

And writing a thesis can be a lonely road for a person accustomed to constant supervision. That's why I think a doctoral degree is proof of project-management skill, regardless of what subject it's in. To get a doctorate, you work on your own for a solid year, meeting with your supervisor about one hour a week.

The key habit that I try to instill is never to let a day go by without progress (except Sunday), and in any case never to let two days go by without progress. Unlike drudgework, research isn't something you can start and stop on a moment's notice. It takes several hours to start up. Usually, once properly started, a worker will have enough momentum to stay productive for several days at a stretch. Then some rest becomes necessary.

With research assistants, the temptation is different. A research assistant is taking courses at the same time. The temptation is to treat the research assignments as "homework" to be done in a hurry the night before the meeting with the supervisor. That, of course, is not what they're being paid for. "Do all work as soon as possible because other people need it" is a workplace attitude that school does not instill.

In either case, if a person's level of effort is inadequate, supervisory meetings become more frequent (but short). Instead of fifteen minutes once a week, you might get ten minutes three times a week, or even five minutes every day. But not for long. I've never had anyone who required daily meetings for more than one week. By then, either they're working productively, or they've proven they're never going to.

(One reason I like really short meetings is that, if you can't say in one or two minutes what you've done, you don't know what you've done, or else it's a laundry list with no unifying purpose.)

A key principle is that unproductive people feel that they are just as "busy" as productive ones, if not more so. Life is full of things to take up your attention. A productive person knows what not to spend time on. I've seen plenty of Ph.D. candidates run aground by keeping themselves too busy with things that aren't their thesis. Don't follow TV shows or sports, or try to be too good at trivial things, or handle too much e-mail... Above all, don't waste time on unimportant things just because they have short deadlines or a fast pace.


Short notes

If you believe the theory that the digits on my birth certificate are in the wrong order, and I was actually born in 1759, then today is my 250th birthday. Hmmm...

I'm too busy to write much; the University is recovering from vandalism that took place over the football weekend, about which I hope to say more later.

Instead, have a look at Greg Mankiw's blog. His point for today: It's rather vague what Obama's health care plan actually is.


More Jupiter

Here's another Jupiter image from the night of the 7th, processed just a bit differently.

120 channels and nothing to watch

Like many other Americans this year, we've decided to stop paying for cable TV. Each of us thought one of the others liked to watch TV, so we kept paying $60 a month for less and less entertainment. No more.

Quite a while back, we started using the Internet, rather than TV and newspapers, for news and weather forecasts. I still listen to local news on the radio because WGAU does a good job of gathering and reporting it. But we certainly don't need TV for the news and weather.

So how will we entertain ourselves? Besides books and music (of course), we have lots of videotapes and DVDs; Melody gets cheap video rentals in the mail from Netflix; and, most importantly, there's lots of TV on the Internet, including a huge archive of TV shows on Hulu. Who needs broadcast TV when you can directly call up a library better than what the TV station has?

Then there's all the music in the world, on Last.fm. And I do mean all. For example, have you ever heard a country song in the style of Hank Williams, sung in Swedish? Click here to have that priceless experience. The button to play it is at the upper right of the page.

And the best thing about it? Never again will I accidentally stumble upon an episode of "Real Housewives." (Neither real, nor house, nor wife. Why do people want to watch obnoxious women cat-fighting?) More generally, the era of "120 channels and nothing to watch" is over. Cable is the wrong way to deliver TV.


Leading economic indicators

I haven't written about the economy in a while. There is, this week, a flurry of indications that good times are just around the corner. Employment will be the last thing to recover; the stock market is recovering now, and the real estate market, even more so.

As I've said, I don't think we'll ever return to the economy we used to have. From here on, lending will be less abundant, and (for a change) lenders will care whether they're going to be paid back. This makes it hard to predict inflation. Various government stimulus measures have expanded the money supply, but a semi-permanent drop in lending tends to counteract that.

Biological facetiae

Sharon pointed this out to me:

The subject of that, in case you didn't catch it, was lab chemicals for DNA research.


A clumsy start with eHow

Last night (Sept. 9) I tried to join eHow in order to publish "how to" articles there and earn royalties on them.

I signed up and uploaded a version of the Linksys print server article that previously appeared here. (Republishing your own material on eHow is explicitly permitted, if you're the author.)

Their computer sent me an e-mail accusing me of plagiarism and saying the incident was being investigated.

Then I got 75 "friend" requests from people I had never heard of; eHow has a social network (which it needs as much as a fish needs a bicycle), and a certain number of people are asking me to give high ratings to their articles, and in return they'll give high ratings to mine. Others are running the Nigerian scam.

[Update:] All this is sorted out now, and you can read my article here. It doesn't cost anything to read eHow; it's advertiser-supported.


Unhappy with Microsoft

This afternoon (Sept. 9) my computer spent a solid hour installing a Windows update, and then it developed the endless-reboot problem just like last month, and I spent another hour fixing it.

If I hadn't been a computer professional, this would have been a case of, "Automatic update broke my computer, and I had to pay a technician $100 to fix it," or something like that. I'm not happy with Microsoft.


Lunar scenery

On the evening of the 7th, I also got good pictures of the lunar crater Tycho (first picture) and the walled plain Janssen (second picture), using the same telescope at lower magnification. These aren't first-rate images, but they gave me good practice doing multi-point alignment in RegiStax.



Jupiter is still up there, and late on the evening of the 7th, I got a rather good picture of it with my 8-inch telescope in my driveway. As usual, I used a modified webcam to record about 2000 frames of video, then stacked and processed them with RegiStax 5.

[Update:] I accidentally posted a mirror image, flipped left to right; that has been corrected. As is customary in astronomy, the image has south up. I've also added a second version processed for higher contrast. But look at Brian Combs' picture taken with a 14-inch telescope at about the same time!



When I was on the radio, back in 1997

On January 3, 1997, radio station WGAU interviewed me on the subject of "The Future" as an episode of the Dick Mendenhall Show. By permission of WGAU, you can listen to that 47-minute interview here, if that's how you'd like to spend 47 minutes. You can even download the interview as an MP3 file.

Click here for a 5-megabyte download with somewhat lower-quality audio, or here for a higher-quality version that comprises 16.5 megabytes. [Actually 14 MB now that I re-saved it with slightly different options.] Both versions have been digitally denoised, and the higher-quality version probably sounds better than what actually went out on the air.

Check out the continuing updates to my recent entry about the Linksys print server. Linksys says it only supports Windows XP and Vista 32-bit, but I've made it work with Windows 7, Vista 64-bit, and even Linux. (The Linux solution should generalize to other operating systems.)


Review of xFountainPens.com X750 "Legislatur"
A fine pen at a surprisingly low price

The people at xFountainPens.com were kind enough to send me a pen to try out, namely the X750 "Legislatur" (their pens, though made in China, have German names).

My impression: Solidly built and easy to use. Compared to other fountain pens, this one is a bit more ruggedly built (which is good; beginners won't have to worry about damaging it). At the same time, it is well finished and has a generous ink flow, so experienced fountain pen users will be able to use it immediately. I didn't have to do any cleaning or tuning to get it to work well.

The pen arrived with a converter to enable filling from an ink bottle, but it also accepts regular and "long" international-standard cartridges. At the moment I have a long Pelikan 4001 cartridge in it.

The nib is a bit larger than on other pens. The pen came with an 18K-gold-plated nib that is silver-colored; a gold-colored version is also available, but no other widths. I don't know what kind of metal it's made of.

As you see in the second photo above, the build quality of the "Bülow" nib is excellent. I have seen worse on $300 pens. Bear in mind you're looking at a pen with ink in it, so there are some specks where droplets of ink have dried; the metal itself is smooth.

The point is 0.84 mm wide, a bit wider than other people's "medium" width. The line that you're writing widens slightly with down pressure, as with Pelikan pens, yet because of its large size, the nib works well over a wide range of pressures.

This would be an economical pen if priced at $50, but the actual selling price is much lower. I highly recommend it.


Linksys WPSM54G or PSUS4 setup without setup disk
(for Windows 7, Vista 64, XP, Linux)

This item has been moved to eHow. Please click through and read it there.


KB972036 — an ill-fated Windows update

My office computer developed the endless reboot problem again today (Sept. 4), and I recovered from it the same way. The update that was trying to install was the new version of KB972036, which I then downloaded and installed manually without incident. There was a buggy version of KB972036 two or three weeks ago, and I'll bet that's what caused the problem the previous time I had it.

I'm not happy about having lost two afternoons of computer use because of errors on Microsoft's part.

Tomorrow I'll have a review of an xFountainPens.com pen. Here, as a sneak preview, is a microscope photo of the tip. Solid and smooth, isn't it?


Two short notes

I've received a fountain pen from xFountainPens.com and it looks really good, with a very solidly made, smooth nib. Full review soon.

If you get a robocall (a recorded advertisement by telephone), please fill out the FTC's online complaint form. Robocalls from commercial businesses are now illegal even if you're not on the do-not-call list.

I hadn't gotten one for a long time, but as soon as the new law took effect, it only took three days for one of them to call me, complete with caller ID data (probably fake). Some people just can't stand not to be in trouble.

Or maybe they're testing possible loopholes. The funeral insurance vendor that called me probably claims to be a charity or something else exempt. It reminds me of the jeweler who called the whole town a few years ago and claimed his calls weren't advertisements because he was just inviting people to come see the new store, or something like that. Things like this go over very badly when they ring, in succession, all the phones in a hospital or dormitory.

[Update:] Insurers are not regulated by the FTC and are thus exempt from the current regulations, unfortunately.


Answers to conundrums

See the partial answers added to yesterday's entry.


Two conundrums

Comments added the following day.

(1) There's lots of good economic news, so the stock market is going down. Why? Because in September the stock market is supposed to go down. At least, that's what I read in the papers.

It was a rumor that a major bank was about to fail.

(2) I'm puzzled as to how to update the window in WPF. In WinForms, I can run a computation that periodically changes a label or appends text to a textbox, and the changes show up more or less immediately. In WPF you don't see the changes until the computation is complete and the program is again listening for events.

The usual Invalidate or Update methods don't seem to affect this. I found a very obscure piece of code here that does force the UI to update. It's reproduced below, but I don't understand it. Also, various cognoscenti are telling me all long computations are supposed to run in separate threads and tell WPF when to update by raising events, but I haven't found the details. There needs to be a simple way to make WPF update the screen whenever there is new data to be displayed.

The essence of it is apparently that WPF is multithreaded — the updating of the screen runs on a different thread than your computations. The code below tells the updater to perform a trivial action, and to perform it with high enough priority that other pending requests are also handled. You're expected to create yet another thread when running a long computation, and I have not yet sorted out the best way to update the screen while doing so.

public static void ForceUIToUpdate()
    DispatcherFrame frame = new DispatcherFrame();
        new DispatcherOperationCallback(
            delegate(object parameter)
                    frame.Continue = false;
                    return null;
            ), null);


Short notes

Why didn't I notice this before? Office 2007 can "save as" Adobe PDF, even if you don't have Adobe Acrobat installed. There was reportedly quite a patent dispute about this. Apparently Adobe considered PDF to be an open standard until Microsoft started using it.

Enforcement by computer: I'm trying to sell my PCMCIA (PC Card) Sound Blaster (for high-quality sound on laptops) via Amazon (where I already have some used books for sale). In the first version of the listing, I made the mistake of saying, "...download updated drivers from creative.com." That got my ad blacklisted because whenever their computer finds part of a web address in a listing, it assumes we're trying to divert sales away from Amazon. It then kept rejecting revised versions of the ad that didn't contain the address. What was annoying is that no human being can be contacted about this. This is 1960s management-by-computer at its worst.

(Click here now — the problem has been fixed and the listing is active.)

Another theory about what's wrong with the public schools: Management guru Peter F. Drucker often pointed out that one of the biggest developments of the twentieth century was the art of running large organizations with shared expertise, i.e., multiple layers of managers who have real knowledge and get to make decisions, rather than a "command-and-control" method where a few people at the top make all the decisions and simply give orders to carry them out. Well — based on long conversations with public school teachers whom I won't name — I think part of what's wrong is many public school systems are all "command-and-control" where people at the top don't trust their subordinates and aren't trusted by their superiors.

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