Michael A. Covington    Michael A. Covington, Ph.D.
Books by Michael Covington
Consulting Services
Previous months
About this notebook
Search site or Web

Daily Notebook

Popular topics on this page:
Blue moon — or is the moon colorless?
Wrong Windows version or wrong CPU?
Review of DeLite 18.2mm eyepiece
Two moons on August 27? No...
When you see something strange in the sky
Exploding IC without known electrical overload
Customizing the Windows 10 Start Menu
Programs that arrive as .exe without an installer
What does "What does your name mean?" mean?
Mars, the moon, and August 27
C#, WPF: Write in a TextBox and force a screen update
Nuts and bolts
Crimp connectors
Moon (Mare Humorum)
Moon (Schickard)
Moon (Aristarchus)
M17 (Omega Nebula)
Many more...
This web site is protected by copyright law.
Reusing pictures or text requires permission from the author.

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

For the latest edition of this page at any time, use this link: www.covingtoninnovations.com/michael/blog

The ads that you see here are not controlled or endorsed by Covington Innovations. They come from Google based on your browsing history.
If offensive ads appear, click on "AdChoices" and then "Ads Settings" and make changes, and/or turn off cookies in your browser.

Click here to see books and other items (mostly astronomy-related) that I am selling through Amazon. More will be added in coming days.


Can you use 14-gauge stranded THHN house wire in a car?

Here's another engineering triviality that is interesting because of what it shows us about how people think.

A few years ago, I got a bargain on two 500-foot rolls of high-grade stranded copper wire, the kind used in conduits for house wiring and industrial machinery. I've used it more than once to install ham radio equipment in cars, and it held up fine. But I got curious and searched the Internet as to whether there was any recommendation not to use this kind of wire in cars.

This will interest others because the same wire is available (in several colors) at Lowe's and Home Depot and is often sold by the foot.

The wire in question is 14-gauge stranded. Let me emphasize that the copper part is just like 14-gauge automotive wire. The only question has to do with the insulation.

The inscription printed on the wire is:


The curious use of "or" is because "MTW," "THWN-2", "THHN," "GASOLINE AND OIL RESISTANT II," and "AWM" are all names of specifications.

So what are we looking at? Wire with PVC insulation and a nylon overcoat. THHN stands for Thermoplastic High Heat-resistant Nylon-coated. The other specifications tell me the wire is rated for temperatures up to 90 C and is flame-resistant. It's insulated for 600 volts.

Automotive primary wire, which I would have used otherwise, is commonly rated for 80 C and is also insulated for 600 volts. There is a high-temperature version rated for 125 C. So it looks like what I have is in between the two, and slightly tougher in other ways.

So... Is the house wire safe to use in cars? I searched the Internet and got a bewildering array of answers.

Some were obviously talking about other kinds of THHN wire. In particular, solid wire (not made of multiple strands) is not suitable for automotive use because it doesn't hold up as well to vibration. No controversy there. And some kinds of stranded THHN wire in larger sizes apparently have fewer, larger strands than automotive wire ought to.

Many of the other answers reflected cultural and historical rather than physical facts. The automotive industry has its own standards for wire, which have evolved separately from those in the electric power industry. That doesn't mean there's any significant difference in the wire, only that it went through different approval processes.

A few people reasoned simplistically: "You have to have automotive wire for automotive wiring, house wire for house wiring, speaker wire for speaker wiring..." Now, I won't argue with the "house wire for house wiring" part. Safety and fire insurance regulations are very strict. I would never do any house wiring with anything not officially approved for the purpose, for reasons of fire safety.

My car, however, is not a house. A big difference bewteen automotive wiring and house wiring is that the automotive wires can be inspected at any time, along their entire length, and will be seen often. Also, I will have full control over where they go. And I know this one will never carry more than 10 amps, though rated for 15.

The other big difference is that there is no official wiring code for cars. I won't void my insurance by using unapproved wire. There are many different kinds of wire in the car already, many of them obviously less tough than this.

So... is there any physical reason not to use this kind of wire? I don't see any. As we saw, the copper conductors are exactly right. It's gasoline and oil resistant. It is rated for higher temperatures than ordinary automotive wire (I looked up the specifications). And, because it's marketed for house wiring subject to building codes, I have confidence that it actually meets the published specifications. Cheap automotive wire, imported from the Far East, inspires a bit less trust.

I'm going to use it. This wire will be well away from the engine and will not be subjected to extreme temperatures or vibration. I don't think the differences between varieties of 18-gauge stranded wire make a difference. Certainly not the mere question of whether it has the word "automotive" on the packaging.

How to change the color of wire

If wire isn't the color that you need, you can spray-paint it. About 20 years ago I did this to get elegant red, green, and blue coaxial cables in a prototype of a piece of video equipment.

If all you need is to mark the wire with a distinctive stripe, you can run a magic marker along it. I've done this many times.


The art of the crimp connection

Getting ready to do automotive wiring, I've been educating myself about crimp connectors, which (believe it or not) are widely discussed on the Internet. (See especially this poster from Molex and this very good video tutorial.) Here are some gleanings...

(1) In high-vibration environments, crimp connections can be more reliable than soldered connections. Solder is stiff and brittle.

(2) There are two kinds of crimp connectors, open-barrel and closed-barrel. The closed-barrel ones are cylinders into which you insert the wire; more about those shortly. Open-barrel connectors, used in electronics but not much in automotive work, are U-shaped in cross-section; you crimp one by bringing down another U shape onto it, upside down. Many of them have two U's, one to grip the wire and the other to grip the insulation, farther back. They're shown on the left side of the Molex poster I linked to.

(3) There are two ways to crimp a closed-barrel connector, a confined crimp and an indented crimp. Confined crimps are better, if done well. Indented crimps are better if you need to get good results with a tool that is not an exact fit. More about tools shortly. Most importantly, the tool matters; don't use pliers unless it is an urgent situation and you don't have anything better.

(4) Crimp connectors are for stranded wire only, not solid wire.

(5) You may need to twist the wire a little to gather the strands together, but don't twist it a lot; it could then loosen by untwisting after you've crimped it.

(6) Can you put two wires into one barrel? This is debated, but apparently the answer is yes if you're careful. Doubling a wire makes it 3 sizes larger in American Wire Gauge; for example, two 18-gauge wires together equal one 15-gauge.

When putting two wires into one connector, first join the wires together with tape or a zip-tie maybe half an inch away from where the connector will go, so that they will go in parallel. Then, don't twist them together; just insert them side-by-side or one on top of the other. Finally, make a good indented crimp manually; what you're crimping is not going to be an exact match to a ratcheted crimper.

The following pictures show some crimping tools. The first one is a versatile hand tool that I think I bought from Radio Shack a few years ago. With care, it handles a great variety of connectors. When making an indented crimp, put the indentation in the seam of the barrel (where there's a tiny gap in the metal), which is normally the top of the connector. If you are going to have only one crimping tool, get this kind.

The second one is my new Tool Aid 18900 ratcheting crimper. The jaws ratchet together so that you can insert the connector and wire, clamp it down, and then use two hands to apply plenty of pressure. When squeezed to the specified position, the ratchet suddenly releases. This is the best way to make good connections quickly if your wires and connectors fit it perfectly. This one fits standard color-coded insulated automotive connectors (red, blue, and yellow, including the newer translucent ones) with wires of the correct sizes in them. Tool Aid also makes a similar tool with interchangeable dies for other kinds of connectors.

Since this tool came from Tool Aid with no documentation, I've tried to be thorough. The ratchet adjustment is likely to be correct as it comes from the factory; adjust it later, as wear has taken its toll and the springs have gotten weaker. Always insert connectors so that the wire is on the side with the color-code dot; the connector gets crimped in two places, once for the metal barrel and once for the insulation, which is somewhat crimp-able and helps to provide strain relief. You can see this in the picture at the top of this article.

One additional note. With a ratcheting crimper, make sure the connector and wire are properly positioned before you squeeze the handle. If the tool has started to ratchet and the connector isn't straight, use the release lever and start over.


C#, WPF: Write in a TextBox from any thread and force a screen update

Here's a handy bit of C# code:

/// <summary>
/// Write a line in textBoxLog and force an immediate screen update
/// </summary>
void Log(string s)
	// If necessary, hand this method call over to the thread that owns textBoxLog
	if (!textBoxLog.Dispatcher.CheckAccess())
			textBoxLog.Dispatcher.Invoke(() => this.Log(s));

	textBoxLog.Text += s;

	// Force dispatcher to empty its queue
	textBoxLog.Dispatcher.Invoke(() => { }, System.Windows.Threading.DispatcherPriority.Render);

This allows you to write in a TextBox from any thread (not just the one that owns the box) and then forces an immediate screen update. You could make the TextBox another argument to the method.

In fact, the last line alone is worth remembering — it provides a way to force a screen update in WPF.

The nuts and bolts of nuts and bolts

I've been filling gaps in my knowledge of mechanical engineering, and here's a practical point I came across: How tightly should nuts be tightened on bolts? In fact, why do we even get to ask this question? Why doesn't the nut stop abruptly when it reaches the end of its travel? Why does it turn a little further if you apply force?

The answer, of course, is that something is stretching. First (although handbooks don't mention it) surely the threads themselves are bending a little. Due to manufacturing imperfections, they weren't tightly in contact, all the way around on every thread, until some pressure was applied. But this is only the first step. What happens next is that the bolt is stretching. That's what will keep the nut from coming off — the bolt is acting like a spring, pulling it tight.

The stretch is only slight, of course, but it's real. And you can break a bolt by just tightening a nut on it if you use too much force. (This is an experience every budding mechanic should have, under controlled conditions, in order to know what it feels like.) More seriously, if a particular bolt can hold 1000 pounds, 800 of those pounds could be the tension on the nut — and that's as it should be! The tension needs to be appreciably greater than any variable external load so that the nut doesn't loosen.

When you tighten a nut properly "by feel," you are preloading it just the right amount, to where the bolt has stretched enough to keep the nut tight under load. More about how to do that here. One recommendation is to break a bolt experimentally, then use half that much torque to tighten it in normal use.

The tensile strength of common machine screws (6-32 and larger) is in the hundreds of pounds, or more. (Chart here; more discussion here.) Automotive-grade bolts are much stronger, and it does no harm to choose a bolt with much more strength than you calculate is needed.

Torque wrenches are only slightly more accurate than a reasonably experienced mechanic's sense of touch, so they are used only in critical situations. Beware of power screwdrivers that apply too much torque — a battery-operated impact driver can deliver 100 foot-pounds, but the recommended torque for a 1/4-inch bolt is about 4. Of course, to some extent you can judge this by how much torque is applied to your hand when the bolt or nut stops turning; let go of the trigger at the right time.

Lubrication on the threads makes it much easier to tighten the bolt and nut. Less torque will then produce more tension. Any oil helps (which is probably one reason why, in older machine shops, almost all metal parts had a thin film of oil on them), but moly grease is said to be especially effective. It also keeps threads from seizing (cold-welding), which is a problem with large bolts carrying heavy loads.

What is the difference between a screw and a bolt? A century or two ago, a screw was any threaded cylindrical object, and a bolt was any metal rod that held things together, whether threaded or just bent at the ends (cf. the "bolt" on a door).

In modern usage, a bolt is tightened by turning the nut, and a screw is tightened by turning the head. It follows that Phillips-head machine screws are both screws and bolts. Carriage bolts, whose heads cannot be gripped, are purely bolts, and wood screws, which do not take nuts, are purely screws. Now you know. But this is only approximately true. Bolts are still called bolts if they go into fixed threaded holes, such as those that hold the head on an engine.


One moon on August 27

I have a pent-up supply of things to write about in the Daily Notebook, so stand by for a flurry of material. But first, photographic proof that, hoax or no hoax, there was only one moon in the sky on the night of August 27:

Canon 60Da, Sigma 105/2.8 lens, handheld, digitally sharpened.


On several recent evenings I've done lunar photography with the Canon 60Da in "movie crop" mode (full-resolution video), stacking 3500 to 7000 frames using AutoStakkert. I'm satisfied that the Canon's image compression does not harm the results; whatever I'm losing, I'm making up by stacking such a large number of frames. These are all with the C8 EdgeHD at the f/10 Cassegrain focus, in somewhat unsteady air, processed with PIPP, AutoStakkert, RegiStax 6, and Photoshop.

Here's Mare Humorum, with some interesting rilles in the crater Gassendi at the top:

Here's the great crater Schickard:

And here's the "spooky" crater Aristarchus, site of occasional unusual reflections of light and perhaps even gas emissions, and surrounded by unusual topography:

An unusual stacking artifact

While processing the image of Mare Humorum, I came across some artifacts that looked like scratches on film, along with square patches (not so visible here). I attributed them to Canon's image compression. And I was wrong.

Those were in fact a problem with AutoStakkert, which was having some trouble doing multipoint alignment. I re-did it with fewer alignment points, farther apart, and the problem disappeared.


How gossip works: Mars, the moon, and August 27

Tomorrow is August 27, and people are circulating messages (especially on Facebook) saying there will be two moons in the sky, or Mars will look as big as the moon, or something like that, and it won't happen again for 200 years. These are accompanied by pictures of the event.

How they got a photograph of something that hasn't happened yet, they don't say.

But the fact is that Mars and the moon won't be doing anything unusual on August 27, 2015. The moon will be nearly full and easily visible in the evening sky. Mars is over on the other side of the Solar System, not easy to see.

The August 27 Mars Hoax started in 2003. My recollection is that Mars did come unusually close to the earth on August 27 of that year (though not a lot closer than it comes every couple of years) and a news release said it would look as big in a 75-power telescope as the moon does without a telescope. If those are not the exact particulars, they are close.

Immediately, people started circulating distorted versions of the story. It has grown every year, until now we're being offered, among other things, two moons.

Why do people believe these things? I think we're looking at an important lesson in how gossip works.

Key point: Some people will believe anything if the message is passed on by a trusted friend. Never mind whether the trusted friend got it or whether the trusted friend did any checking. Back in my computer security days, I once had a rather angry phone conversation with a woman who was circulating a false charity appeal, and her defense was, "My sister-in-law says this is true, and she don't lie!"

To such a person, facts don't matter; all responsibility is shifted to someone else.

Other people will believe anything they read, without inquiring who wrote it. The news release must be from NASA because it says it's from NASA. It never occurs to them that such a claim could be false. Look at the amount of health-related spam I get that falsely claims to be from Harvard or the Mayo Clinic or, curiously, Rachael Ray.

To them I say: The letters on your screen look just as neat whether or not they're telling the truth. This warning is especially for older adults who grew up in the era when only large organizations could afford good printing. Some people are primed to believe anything that is well-written and neatly printed.

A third factor is that some people will believe anything about the sky because they've never looked at the sky and have apparently forgotten those fifth-grade science lessons. How could there be two moons in the sky? That's what I would want to know. If a rare thing is going to happen, please tell us what it is and how it works.

Finally, there are deliberate hoaxers. They aren't the source of all distorted messages, but when Photoshopped pictures start showing up, someone is clearly creating them deliberately. While I don't entirely approve of hoaxers, we have to admit that some of these things may originate in contexts where they are clearly jokes, and other people take them too literally.

One last thing I need to remind people of: Other people will see what's on the Internet even if you don't click "share." You do not need to single-handedly "share" or forward everything that comes your way!

In fact, please don't put your name on anything if you can't take responsibility for it. That applies not only to silly hoaxes, but also to other pass-it-on messages, such as those accusing individuals of animal cruelty or other crimes. We've had a spate of those lately, too.


What does "What does your name mean?" mean?

There's a fair bit of nonsense circulating on the Internet about "what your name really means." As Thomas Parham pointed out on Facebook, not every female name can actually mean "warrior princess," despite what some people say.

If you want to know the real history of a name, let me recommend behindthename.com.

But first, let's get clear what we're asking for. If you just want to make up something mystical and say that your name "means" it, then count me out; I'm not playing that game. Words don't mean things by themselves apart from being understood and used intentionally by people.

Another game I won't play is untestable speculation. Like any other history, the history of words and languages is a realm of facts, not fiction or mere guesswork. Some people don't seem to have caught on to this and will offer you made-up "origins" of words that haven't been checked out. Me? I don't want to make up a story just for the sake of having a story.

Getting back to the meaning of names, the first point to note is that a name doesn't have to "mean" anything — its purpose is to identify a person or place. There's a whole book about this by my undergraduate major professor (John Algeo) as well as a classic philosophical study by Saul Kripke.

Some names do mean things. Consider my wife's name, Melody. The word means "tune" in English, and her parents surely had this in mind when they named her. So this is a name with a definite meaning.

Consider now my older daughter's name, Catherine. This is derived from the ancient Greek word katharos 'pure'. Most people who name their babies Catherine today are not aware of this; they just know it's a nice-sounding name. But at one time it was given to people because their parents recognized it as dervied from the word for 'pure' in their language. (In Greece, they surely still do.) So this is a name that (in the non-Greek-speaking world) used to have a meaning, whether or not the meaning is still remembered.

By contrast, consider my younger daughter's name, Sharon. This is an ancient Hebrew name — and that's all we know for certain. It refers to a plain in Israel and may mean "plain" or "field" or may not; linguists are not quite sure. It was not a widely used word. So this is a name that probably had a meaning in the past, but the meaning has been lost.

Finally, consider the name of the comic-strip character Dagwood Bumstead. Here Bumstead is a real (rare) name (whose origin I do not know) but as best I can determine, Dagwood was made up to sound like an English name while being completely novel. The person who made it up did not attach any meaning to it, as far as is known. This is a name with no meaning — it's just a name, made by rearranging syllables from other names.


Friend or foe? Or do we get to use our brains?

One of the most frustrating things about online forums, and many other kinds of discussion, is that some people are more interested in classifying you as friend or foe (and then combatting you if you are the latter) than in exchanging knowledge.

It's the instinct that underlies tribal warfare. Everybody is either in your tribe or someone you're at war with. Combine that with postmodernism ("there is no objective truth") and you get a mess. There is no information to discuss. Rational discourse is replaced by a mere power struggle.

A slight variation on this theme is the way some people decide that if you don't toe their party line 100%, you must be part of a vast evil conspiracy. To such a person, thinking for yourself is forbidden; you're not even supposed to understand the other side (and it terrifies them if they see you trying to). You're "enemy" if you don't hold exactly the approved set of opinions.

And either of these is an excuse for turning off your brain, refusing to learn anything, and merely dispensing pre-planned standard refutations.

That's not me. I insist on understanding the other side if at all possible. I'm not afraid of finding out what goes on in other people's heads, even if they're people outside my own tribe.

Several times recently, when I asked a question about a controversial issue, people have jumped on me as "one of them" and tried to beat me down with rebuttal that didn't address what I actually said.

My political views are on the conservative side, but I'm fairly often mistaken for a liberal. On the other hand, in a recent discussion of socialism I was taken to be a hostile reactionary. It was strange when someone took my use of the word "socialism" as an insult while advocating socialism. To me, that word is not an insult; it is the name of a somewhat multi-faceted politico-economic movement, which I do not support, but which I don't consider contemptible.

In fact, that's the point. I don't feel contempt for people whose opinions are different from my own. I think they're mistaken, but I don't despise them. I'm not looking for enemies.

Short notes

Gadget of the day: DeWalt right-angle drill-driver adapter. This, with some hex-socket drill bits, enabled me to drill holes in the metal bumper of my car without removing the plastic bumper cover. (Did you know that when it looks like you have a plastic bumper, you actually have a metal one, a few centimeters behind it?) The holes are to hold running lights. More about this soon.

Business is booming and I'm limited only by the amount of time I can spend working. Notebook entries will continue to be a bit scarce for a while.


Unusual failure mode of an NE555P IC

A couple of years ago, I built myself a variable-brightness 12-volt floodlight to use at observing sites. It has a magnetic mount, and when I want to drive away without disturbing the other astronomers, I can open the sunroof of my Ford Escape, attach this to the roof aimed forward, plug it into the lighter socket, and have only as much light as I need — aimed only forward, not in other directions.

The other night, I plugged it in, turned it on, heard a pop, and it wouldn't light up. Thinking I had blown a fuse or burned out the bulb, I set it aside until today (August 15).

It wasn't the bulb or the fuse, so I opened it up. My circuit uses an NE555P (555 timer) IC as a variable-duty-cycle oscillator controlling the gate of a power MOSFET.

And the 555 seemed to be sitting slightly crooked in its socket. Hmmm...

And when I pulled it out, here's what I got:

I'll seek expert advice, but it doesn't look to me as if the chip is burned up. (It's all there, although there's some black plastic on top of parts of it.) Nor could I induce any failures after replacing the 555. The light controller works fine, and neither the 555 nor the MOSFET gets warm.

Was this 555 defective from the factory? Imperfectly sealed, and then, after months of being carried around in the car, it broke open?

And if nothing is burned up, what was the "pop" that I thought I heard? Maybe just the sound of the switch? Or did a droplet of moisture get inside the 555, bridge the power supply, and then vaporize and blow the IC open when I applied power?

File this as Exploding IC without known electrical overload.

Customizing the Windows 10 Start Menu

Customizing the Windows 10 Start Menu is something you must do. It starts out looking rather frivolous, like this:

Here's what mine looked like after a few minutes of arranging. Of course, I have a lot of software, and now the Start Menu helps me get to it.

First, a philosophical question: What is the Desktop for? Since Windows 95, it has been somewhat unclear whether the Desktop is for files or for software shortcuts. Of course, it's useful to be able to mix the two. And software manufacturers like to put links on the Desktop so it will be obvious their stuff is on your computer.

Through Windows 95, 98, 2000, XP, 7, and 8, I didn't really resolve the question. The Desktop was for everything and gradually got more cluttered until I periodically cleaned it up.

Under Windows 10, I've decided that the Desktop is for files and folders I'm currently working on, along with shortcuts to software that is intimately tied to the current project but not otherwise often used. The Documents folder is for files and folders I'm no longer actively working on. And where do I put software? The Start Menu, nice and big, is for frequently used software.

Now then. The first thing to notice about the Start Menu is the All Apps button at the lower left corner. Click it, and you see all your software, alphabetically and in groups, just like the Windows 7-and-earlier Start Menu.

So rest assured, you can take software off the Start Menu and still get to it.

If you want the Windows 7 experience, you can take everything off the Start Menu and just use All Apps. But who wants the Windows 7 experience?

OK then. The first thing I did with the Start Menu was stretch it up and rightward to make it really big. I like the big Start Menu of Windows 8, so I made this one as big as possible, too.

The next step was to unpin (delete) all the software I don't actually use. Right-click on an item and click Unpin From Start, and it's gone (but still exists in All Apps).

Next after that, create and arrange groups, as follows:

  • To move an item into a group, drag it so that it partly overlaps what's already in the group. You'll quickly get the hang of how to move items around.
  • Every group has a title bar at the top, whether or not you can see it. (If it has no name, there's nothing in the title bar.) Mouse over the title bar and click the "=" symbol at the right, and you can name or rename it. Move groups around by dragging by the title bar.
  • A group automatically vanishes when there is nothing in it. Unpin all the items or move them to other groups, and the group is gone.
  • To create a new group, drag an item into the empty space at the bottom right of the Start Menu. It becomes a group all its own.

What about software that doesn't show up on the Start Menu? Easy. Find the piece of software in All Apps, or by navigating to the .exe file, or finding any shortcut to it; right-click on any of these; and choose Pin To Start. The software will appear at the lower right of the Start Menu, in a group all its own. From there, move it into the group where you want it.

One last heuristic. The Start Menu is only for software you use often. If you use something every six months, it's fine to look for it in All Apps when you need it. The Start Menu is to speed you up, and I put the most commonly used items at the top.

There's more. You can resize any icon on the Start Menu to make it a quarter its normal size. You can customize what's in the "Most Used" menu at the left. And so on. I haven't explored all these yet.

Software that arrives as an .exe without an installer

What do you do with software that arrives as an .exe file, or as a folder containing an .exe plus other files, but no Windows installer?

The instructions usually say, "Put this in any convenient folder, and click on the .exe file to run the program."

For a long time, I tended to put these on my desktop. But under Windows 8, I came up with a better way.

What I did was create a folder called Programs under Documents. (Thus, C:\Users\Michael\Documents\Programs.) In it I place each of those executables, or the folder containing the executable plus its accompanying files, as appropriate.

Then I right-click on the .exe file and choose "Pin To Start." That puts it on the Start Menu, and I can move it into an appropriate group.


R.I.P. Randy Glasbergen

Cartoonist Randy Glasbergen, whose work I've been enjoying every morning for years, has died. I noticed that his web site had stopped updating; I hope it will be kept up as a memorial. He was noted for his clean sense of humor.

Goodhart's Law

I've learned the name of a management principle that applies to education and other activities. Goodhart's Law says that a measurement loses its value as a measurement when it becomes a goal.

This applies to moderately complex activities, of course. If the activity is simple enough, then adopting a measurement as a goal (such as maintaining the level of water in a lake) makes sense.

In education, achievement tests are a useful measurement if they are not the overall goal. But as soon as test scores become the only goal, two things happen: teachers start teaching to the test, and the test scores cease to be a measure of real educational achievement. You have a small feedback loop that doesn't include all the things it was intended to.

More generally, you're managing only what you can measure, not what you really need to be achieving.

Goodhart's Law originated in economics, where its applicability is obvious. Once you adopt a goal of, say, having a particular inflation rate, you will find that policymakers and investors can alter the the inflation rate by doing things you never planned on, and it no longer tells you much about the state of the economy — that information has been pushed into other variables that are not being manipulated.

The same goes for any other kind of management where a measurement is substituted for the real goal. I leave it to the reader to think of other examples.


Cleaning up the new computer

It was time to remove the bloatware that came with my new computer. I know it's well-intentioned, but I don't need another antivirus program and another backup utility.

So, after making a set of installation DVDs, I removed Dell Backup and Restore, various other Dell things, and the McAfee antivirus suite.

All of this proceeded smoothly from Control Panel, Programs and Features, but removing the Dell things took more than 5 minutes with only a barber-pole-striped progress bar showing. I thought it had gone dead.

And of course McAfee warned me sternly that everything I hold dear is about to be eaten by computer viruses. Translation: They want me to sign up for a subscription.

I don't recommend removing software of this type unless you know what you're doing. But if the preinstalled software nags you, slows down the computer, or asks for money (for a subscription that you don't feel you need), it needs to go. I already trust Microsoft's Windows Defender, which is built into Windows 10 and doesn't need to be fighting with another piece of software trying to do the same thing. And I make my own backups with Acronis and with my own cloud storage system.

The computer runs faster now that it's unburdened!

The next big thing: customizing the Windows 10 Start Menu. As with Windows 8, this is necessary and I'm afraid many will be put off by the frivolous stuff that Microsoft positions so prominently until you move it aside. More about this later.


Holey wall, Batman!

I've recently had my consciousness raised about how to fix holes in walls. In particular, I've become aware of the difference between spackle and joint compound ("sheetrock mud"). Until now, I had only used spackle.

Spackle is a relatively hard putty and is for filling holes. It dries very hard and needs to be sanded heavily after use, unless the hole was very small.

Joint compound goes on as a soft putty and is for filling dents in surfaces. Unlike spackle, it can be spread very smooth with a wide putty knife and may not need further sanding, depending on how smoothly you applied it. It dries much more slowly, taking about a day.

I'm fixing a wall in our living room where the sheetrock began to delaminate. For whatever reason, there's a fairly large patch where the outer cardboard layer wrinkled and came apart in layers when we took down a supposedly easy-to-remove decorative panel. If there had been any plumbing near it, we would think it had gotten wet.

Spackle didn't work so well because I couldn't get it very smooth, and in sanding it, I had to apply enough lateral force to cause more delamination. I've gone at it again with joint compound. I had to put it on rather thick and will need to sand it and apply a second coat. But I think I'm winning. If not — well, in a few years we will put built-in bookcases there.


In memoriam

I want to take a moment to honor my mother, Hazel Roberts Covington, 1925-1985, who departed this life 30 years ago today, having spent her last two decades making great sacrifices and working very hard to provide a good standard of living and a first-rate education for Julie and me after our father was killed in 1966.

Here you see my parents' wedding portrait, 1946.

What to do when you see a strange object in the sky

Occasionally I get inquiries like this: "I saw something strange in the sky the other night. No, I don't remember exactly where or when, and I can't really tell you what it looked like. Can you tell me what it was?"


If you see something unexplained in the sky and think you might want to ask an astronomer about it, here's what to do.

(1) Make note of the place, date, time, and direction. This is important. What's in the sky depends on the earth's rotation and orbital motion. We need to know where you are in space and time, and what direction you're facing (north, south, east, or west). If you don't know which way is north, at least take note of something like streets or buildings that you can later find on a map.

(2) Find a reference point in the sky. This is important. Find something else in the sky that you can compare your object to. Can you see the moon? Any bright stars? Or even clouds? This is important for establishing what part of the sky you're looking at and whether the object is moving.

(3) Describe it in words, and remember that "big" doesn't mean "bright." Put the description into words while you're still looking at the object. Distinguish brightness from size. A bright star is a tiny pinpoint. A big cloud need not be bright.

(4) Determine size with your fists and fingers. Your brain will try to tell you something false about the size of the object. You may feel that the object is the size of a dinner plate, or an airliner, but those are based on false assumptions about its distance.

What you need to know is how much of the sky it covers. And that's where your hand at arm's length is your key instrument. Can you cover the object with one finger? Is it more like a thumbnail or like a fist? That's what we need to know.

(5) Determine movement, but be careful. Some people see everything "zigzagging around" in the sky. That's because their head is bouncing around and they don't distinguish movement of the object from movement of their head or eyes. (Or a bouncing handheld video camera!) That's where comparison to other objects is vital, even if they are only clouds.

Again, use your arm as an instrument. Point to the object with your finger and take note of how fast your arm is moving.

There. Now you're ready to ask me what you saw.



Two moons on August 27? No...

It must be August Fools' Day. A truly remarkable amount of foolishness has been flung at me through the Internet today.

All of a sudden, this evening (Aug. 5), people are asking me whether there will be two moons in the sky on August 27.

If you asked me about this, let me ask you two questions.

(1) Did you see this on a reputable web site that somebody takes responsibility for? Or just in a "pass in on" message? Have you ever known "pass it on" messages to tell the truth?

(2) Given what you've known about the Solar System since fifth grade or so, do you think it can happen?

The hoax seems to be distantly derived from a (correct) news release that on August 27, 2003 (12 years ago) Mars would look as big in a 75-power telescope as the moon does without a telescope.

There. It is not my job to refute every silly thing everybody thinks of. Fifth grade science should have taken care of this one. Or if not that, then at least Google.

If I sound tired, it's because the day's shenanigans included plenty of other nonsense, including people wanting me to share fake airplane tickets, a personal-harassment campaign that I busted on Facebook, a flood of spam, and even phone calls from charlatans.

Did someone dump out the loony bin?


Dodging clouds at Deerlick

On August 4, I trekked down to the Deerlick Astronomy Village with my Celestron AVX mount, 300-mm Canon telephoto lens, and Canon 60Da, but no autoguider. After doing a careful polar alignment and turning on PEC playback, I got excellent tracking — but had to dodge clouds.

So the trip turned into an equipment-testing run, where I collected lots of proof that this setup can make perfectly-tracked 1-minute exposures without an autoguider. Except for the first two or three minutes after slewing to a new object. The first exposure in each series was poorly tracked because slack in the gears had not yet been taken up. I could probably have prevented this by carefully preloading the mount to make it slightly out of balance, heavy to the east. Instead, I just skipped the first picture in each series.

Most of the pictures were unsatisfactory because high clouds drifted into them, but the tracking was excellent. And I did get one good picture. Here is the full field of M17 (the Omega Nebula) with a bit of M16 visible at the top. Stack of six 1-minute exposures, calibrated with dark frames but no flats (this lens has hardly any vignetting).

To show you how good the tracking was, here's just M17, cropped rather than downsampled. You are looking at the actual pixels of the Canon 60Da.

Not too bad for a cloud-dodging expedition!


The best eyepiece ever made?

My new Tele Vue DeLite 18.2-mm eyepiece arrived on August 3, and I'm happy to report that it works better than any other eyepiece I've ever tried.

The DeLite line is getting rave reviews everywhere, and what is unusual is that hardly anyone ever says that any other eyepiece is better. Usually, an eyepiece review discusses a complex set of trade-offs. Not this time. DeLite seems to be beating the competition in several ways at once.

In fact, the undignified name may be its biggest drawback. It is intended to be a "lightweight" version of the Delos wide-field eyepiece. It has only a 60-degree field, but some reviewers are saying it is sharper and/or has better contrast than the Delos.

Here's my own experience. As I've noted, the EdgeHD telescope is something of a game-changer for eyepieces because it delivers a wide, flat field. Now we can be sure that any blurring at the edges of the field is the fault of the eyepiece, not the telescope. Owners of good refractors have had this advantage for a long time.

Bottom line: You judge an eyepiece by what you don't see — blurs and artifacts — and the DeLite is the most artifact-free eyepiece I've ever tried.

I've compared the 18.2-mm DeLite with a 14-mm Radian and a 21-mm Pentax XL.

I did the first comparison without astigmatism correction, since my astigmatism is not severe and I wanted all the eyepieces to be on a level playing field. The Pentax performed well but had definite field curvature — I couldn't get the edges in focus at the same time as the center. The Radian had a flat field but some edge blur; the central 85% of the field was sharp. The DeLite was similar to the Radian but better; the central 95% of the field was sharp, but there was some degradation right at the edges.

The DeLite also seemed to give a sharper image than the Radian, somehow, and I don't think the lower magnification was the only reason.

I had no problems with eye position with any eyepiece, but the Radian seemed a little pickier about eye position than the others.

Then I added 1.00D DioptRx astigmatism correctors to both the DeLite and the Radian. This enabled me to focus the stars to sharper pinpoints than before — especially with the DeLite. I should explain that the Radian is an excellent eyepiece, but the DeLite seemed a little better in both sharpness and contrast. Aside from improved edge-to-edge sharpness, which is definitely objectively real, the apparent improvement may have been due to lower magnification. A fairer test would be an 18.2-mm DeLite versus an 18-mm Radian.

We've come a long way. In the 1970s we had excellent telescopes and mediocre eyepieces; in the 1990s, mediocre telescopes with much better eyepieces; and today, we've finally gotten both of them good at the same time. Eyepiece design is an old art — the Plössl dates from 1860 — but clearly, progress hasn't stopped.

Update: Some reviewers are now saying that some other premium eyepiece is better than the DeLite in particular circumstances. So it's not exactly the top of the heap. Also, it is priced comparably to eyepieces with a wider field, and some people put a great value on wide field; I don't. Still, the DeLite is the best eyepiece I've tried and one of the best anyone has ever encountered. It's a definite technical advance.


Why won't your old software run?
Wrong Windows version or wrong CPU?

(Expanding on a topic I wrote about briefly last month.)

Every time a new version of Windows comes out, there's wailing and gnashing of teeth. "Oh, no! My software won't run! I'll have to get all new software!"

But the truth is, Windows has been the same internally since Vista and almost the same since Windows 2000. It would be hard to make software that runs under Vista but ceases to run when you switch to Windows 7, 8, or 10. The internal differences between those versions are tiny.

A few days ago I think I've figured out why people are afraid of new Windows versions. There must be a good many people who, somewhere along the way, got a new PC and couldn't install or run their old software.

Now, a little of this can be attributed to klutzy installation routines that ask Windows what version it is, and simply refuse to install if it doesn't tell them what they want to hear. But this can be worked around. You can install in compatibility mode, which basically means making Windows lie about its version number.

More commonly, they've switched from a 32-bit to a 64-bit CPU. Since 32-bit and 64-bit versions of Windows look just alike to the user, and all modern (32-bit) Windows software runs equally well on both, they might never know anything has changed. Except...

Except that DOS and Win16 applications no longer run.

And the reason is that when it's running in 64-bit mode, the CPU only supports 64- and 32-bit software. It does not have "Virtual 8086 mode," needed to host the 16-bit Windows system.

There are other reasons why old software might not run in Windows Vista and later. The biggest is Vista's security. Programs can no longer write anywhere on the disk — Windows limits who can write in which directories. There are other limitations that should not get in the way of well-designed software.

However, these limitations came in with Windows Vista. If a version later than Vista suddenly broke your software, I'm guessing what really happened is that you switched to a 64-bit CPU. To find out whether you're running 32- or 64-bit Windows, right-click on "Computer" or "This PC" and choose "Properties."

If older software is mission-critical, what can you do?

Two things. All 64-bit PC CPUs (Intel Core, etc.) will also run as 32-bit CPUs that act just like Pentiums, but faster. You can perfectly well install the 32-bit version of Windows Vista, 7, 8, or 10 on a brand-new computer. There's no point in having more than 4 GB of RAM in the computer, because in 32-bit mode, you can't access more than that. But the computer will work. If 32-bit Vista met your needs, 32-bit Windows 10 almost certainly will, too.

The other is to use a virtual machine. That is a program on your 64-bit PC that emulates all the functionality of a 32-bit PC. On the emulated PC, you can install any operating system you want.

There are two catches.

One: You need a license for the operating system in the virtual machine. That is probably not a problem; you probably still have the installation disks and product key from a computer that was scrapped long ago. Or if not, you can buy them secondhand. But you can't install a copy of Windows in a virtual machine that you are currently using in a real machine.

Two: A virtual machine can get viruses just as easily as a real one. That's why, if you choose to install a virus-prone OS such as Windows XP, you need to obtain and install third-party antivirus software (Windows Defender is no longer supported on XP). Or don't allow the inner OS access to the Internet.

How do you get a virtual machine? Fortunately, that part is easy. Several versions of Microsoft Windows Virtual PC are free to download and use, but the cognoscenti recommend VMware Player or VirtualBox (the latter free even for commercial use).

A virtual machine is then simply a big file (often in .vhd format) that serves as the inner (emulated) PC's disk drive. You turn it on and off by starting and stopping the virtual machine software.

Each of these software packages comes with add-ins that you can install in the inner operating system to allow it to communicate more fully with the real (outer) computer, sharing its disks, printer, and network. VirtualBox also provides Remote Desktop Protocol access to any operating system (even Linux or DOS). I'm going to try it out soon.


Windows 10 security (updated)

[Major update, August 4.]

Under the hood, Windows 10 impresses me as reliable and secure, but flawed by two prominent blunders which need correcting immediately. The blunders are ethical and legal, not technical. On the technical side, Windows 10 is solid and even has security improvements compared to earlier versions. But two very bad decisions seem to have been made by non-technical management (or maybe by ethically naive technicians) and somehow got past Microsoft's lawyers.

Bad idea number 1 is Wi-Fi Sense, which is supposed to make it easy for you to share access to a Wi-Fi network with your friends. It does — but "friends" might include Facebook friends who are total strangers and who want to use your network for mischief.

Bad idea number 2 is peer-to-peer sharing of Windows updates. This is actually a great idea within a local-area network. It keeps every machine from having to download from Microsoft separately. But Windows also, by default, uses your bandwidth to upload updates to strangers! That is unacceptable for many reasons.

Both of these are easy to turn off (see the specific recommendations below). I expect Microsoft to come to its senses and abolish both of them soon.

So what were they thinking?

I am wondering why Microsoft made these bad decisions. If I had been their lawyer, I would have told them that:

  • Wi-Fi passwords aren't made to be shared. If a network is free for use by everyone, it doesn't have a password in the first place. Wi-Fi Sense probably violates user agreements every time it is used, and it may violate laws against unauthorized password disclosure (such as this one, to take just one example).
  • When done without users' clear consent, peer-to-peer sharing of updates could constitute theft of services. It can degrade their network performance, impose costs on them (some of us pay for bandwidth!), set off alarms in network routers and the like, and violate ISP user agreements.
  • It won't do to hide the consent deep in the fine print. Users expect the Windows 10 terms of service to be very similar to earlier versions. Courts won't enforce strange, adverse provisions that are hidden deep in a contract where people have been led not to expect them.

I do not have technical details, and it's possible neither of these bad decisions is quite as bad as it looks. But I have to call things as I see them.

All I can guess is that the people at Microsoft were caught up in trendiness, thinking entirely too much like a casual smartphone user rather than someone who does serious computing. Someone forgot that Wi-Fi isn't a public resource; those passwords are there for a reason! And someone forgot that bandwidth isn't free and unlimited.

Am I giving up on Microsoft?

No, I'm not giving up on Microsoft. Nor am I going to spread the vague idea that Microsoft is an evil conspiracy. But I am going to hold them accountable for specific blunders.

I do think they have shot themselves in both feet and need to regain the trust of the public.

I don't think we need to be afraid of Microsoft actually doing great harm to the computer-using public. Far too many eyes are on them! Their main victim is themselves.

On the technical side, they're still as good as ever. I still need a Windows platform, and I still think Windows 10 is the best version of Windows. After all, the Windows 10 development effort was directed entirely at correcting the user interface problems of Windows 8, and improving not at adding new functionality. It should be reliable.

And due to its use of object-oriented technology, Windows is more technically advanced than MacOS or Linux, both of which are based on UNIX (1978).

What amuses me about people who bash Microsoft for vague reasons is that they often praise Apple for doing very similar things. If Wi-Fi Sense had turned up in the iOS world, there would be a chorus of people praising it and demanding that Wi-Fi user agreements accommodate it. Apple's customers are like that. Microsoft's customers are a more skeptical crowd.

And maybe that's how it should be. A sign of a healthy democracy is that people mildly dislike their government. Maybe it's also a good sign with dominant software companies, or dominant vendors of any kind.


The first principle of computer security is stop, think, and use common sense. Think about what you do and do not want computers and the Internet to do for you.

The reason for Microsoft Accounts and OneDrive is to synchronize your data between different computers. But Microsoft can't transmit your data without having your data. So if you want more privacy, don't hand your data over to a third party.

Similarly, some Windows 10 software uses the microphone, the camera, and/or the location data that the operating system keeps track of. In the old days, PC software would just use them, without consulting you. Now, at least some of the time, you have to grant permission. That's why you get all those warnings.

Then there's advertising. Microsoft wants to send you ads and other local information (weather, for instance) based on information about your location and preferences. Some people welcome this, and others are uneasy about it.

What I urge you to do is think about whether you need these things done, and if not, turn them off (and live with the consequences of turning them off).

Remember that web browsing is not and never has been a private activity. Web servers have to know where you are in order to send the web pages to you. That is not new with Windows 10.

Are Windows computers particularly vulnerable to viruses and malware? No; since Vista, Windows has been rather hard to tamper with. But Windows computers are numerous, so that's why the virus authors target them, despite the heavy defenses. Almost all viruses now require some kind of consent from the victim, but it's all too easy to get consent because some people think they are supposed to answer "yes" to anything that pops up on the screen.

Stop and think before you say "yes."

Specific recommendations

Here are my specific recommendations for Windows 10. Steps 4, 5, and 6 are the most important ones.

(1) If practical, set up your user account without linking it to a Microsoft Account.

During the installation procedure, when you're invited to create a Microsoft Account, my recollection is that you have to say you want to create one, then bail out of the process.

This automatically disables Wi-Fi Sense and improves security in other ways.

If your user account is already linked to Microsoft Account and you want to unlink it, here's how: Start button, your name, Change account settings, Your account, Sign in with a local account instead. More explanation here.

I have a Microsoft Account; I just don't want to sign into it all the time when I'm using one of my own computers.

(2) Do not store sensitive information in OneDrive, Google Drive, DropBox, or the like.

You don't know who's going to get hold of it. While I do not know of any breaches, the fact is, you're giving the information to someone else to hold for you. No doubt they will safeguard it in good faith, but they could get hacked or burglarized...

(3) If you are not using Wi-Fi, turn Wi-Fi off.

This is always a good idea. On most laptops and tablets, "airplane mode" neatly turns Wi-Fi off (and also cellular telephony). You also have an easy way to turn off Wi-Fi by itself.

My desktop also has Wi-Fi; yours may, also. I turned it off by using Start button, Settings, Network & Internet, Ethernet, Change adapter options, right-click on the Wi-Fi adapter, Disable.

(4) (IMPORTANT) Make sure Wi-Fi Sense is turned off.

If you have no Wi-Fi, this will not apply. But if you have Wi-Fi, this is important.

Go to Start button, Settings, Network & Internet, Wi-Fi, Manage Wi-Fi settings, and make sure the two switches are off, like this:

(5) (IMPORTANT) Use Windows Defender and Windows Firewall.

Use them in the normal way, and let them send information to Microsoft. That's how antivirus software finds out about viruses. Get to this from Start button, Settings, Update & Security, Windows Defender.

In my opinion, Windows Defender is sufficient; you don't need other antivirus software unless you have unusually hazardous habits (i.e., you enjoy downloading things just to see what they'll do).

(6) (IMPORTANT) Get your updates only from Microsoft and do not share them, except in a LAN.

This is hidden deeply under Start button, Settings, Update & Security, Windows Update, Advanced options, Choose how updates are delivered.

Make sure you uncheck the button that says "...and PCs on the Internet." It's checked by default and will have you sharing updates with total strangers, using bandwidth and risking getting a counterfeit update.

Since I have a local-area network, I use these settings:

If you don't have a LAN, just turn off the blue switch that is labeled "On."

(7) Keep Windows up to date.

I know automatic updates are controversial, but I accept them. I want Microsoft to fix security problems as soon as they are discovered. This is under Start button, Settings, Update & Security, Windows Updates, Advanced Options.

Note that you have the option to defer upgrades. That will delay (by months) the installation of updates that only add features without improving security.

And that brings to mind an important fact about how Windows 10 is designed: There are not going to be numbered versions (10.1, 10.2, etc.). Instead, features will be delivered incrementally through Windows Update. (Who wants to bet that Wi-Fi Sense will disappear soon?)

(8) Go through all the privacy settings and disable everything that does not need to be enabled.

Click Start button, Settings, Privacy. Here are screen shots of some of my own privacy settings:

Because I handle highly confidential data, I've opted not to let Microsoft collect the usual information about my usage.

And for reasons of performance, I've turned off all the background apps, most of which are apps I never use in the first place.

This is not the last word on Windows security, but it should get you started. Keep your ears open for well-informed advice, but steer clear of paranoia.


How to get a heavy telescope onto a mount with a dovetail

One of the puzzles of astronomy is finding the best way to heft a heavy (or even light) telescope and fit it into the dovetail saddle of an equatorial mount. I've found two videos that are useful, if you can tolerate (or even enjoy) the intermittent parody of Arnold Schwarzenegger: Video 1 and Video 2. The second one reveals a clever way to use an adjustable observing stool as a telescope support during the process.

The tube that he is hefting in the demonstration weighs 45 pounds (20 kg).

Putting my 8-inch on the Celestron AVX mount, I find that tilting it in from the side works best; an advantage is that both of the safety screws get to stay on, but the technique would be easier with a wider (Losmandy type D) dovetail.

Another hint: A second dovetail, mounted on top of the tube, not only holds cameras but also serves as a convenient handle!


Blue moon — or is the moon colorless?


Yesterday (July 31) was a blue moon, whatever that is. In brief: The moon didn't turn blue (the picture above is fake). It was merely the second full moon within a calendar month, a rather uncommon event.

I should emphasize that a blue moon is not a physical event. Full moons occur at fixed intervals. Our clumsy calendar has months of varying lengths, not synchronized with the moon. Rarely, the moon is full twice in one calendar month. When that happens, it doesn't look any different from any other full moon.

This is one of several things "blue moon" may mean. It may also mean the fourth full moon in a three-month season, when it occurs; or perhaps, longer ago, a belewe ("betraying," i.e., exceptional) full moon that complicates the calculation of the date of Easter.

So — is the moon colorless? We tend to think so, even to the point that I have, in the past, sometimes decolorized my moon pictures to get rid of a color cast.

But look carefully at the following. This is a straight, unaltered color image of last night's full moon. The picture was digitally sharpened and color balanced for overall neutral color. It was taken with a Canon 60Da (vintage 2012) at the Cassegrain focus of my Celestron 5 (vintage 1980).

Look carefully, and you'll notice that some of the "seas" (dark patches) are browner and some are bluer. These are color differences that you can see with a telescope or binoculars. Here's a saturation-enhanced picture that exaggerates them:

Finally, here's a "close-up" of the Mare Crisium area in rather unsteady air. This was taken by recording about 7000 frames of video with the same camera, but using only the central part of its sensor, and stacking the best 80%, then sharpening with RegiStax. Again, you can see some subtle coloration.

I did an experiment to find out whether I should stack the best 80% or the best 5% (which would still be about 350 frames, enough to give a good stack). The underlying question was whether Registax should pick out the sharpest frames or average nearly all of them (throwing out only the ones that are egregiously bad). The answer? It made no difference that I could see. There was certainly no harm from stacking 80%, and although it wasn't a major consideration here, stacking more frames reduces grain nicely.

Welcome to August...

It's the beginning of August, my least favorite time of year. This is when the University is shut down, the weather is as bad as it gets (very hot, with high clouds precluding astronomy), and this period includes the anniversary of the day my mother died.

Halfway through August, the University suddenly revs up, and so does everything else, and the weather starts getting better. All of a sudden, there's too much to do. So I'm going to slow down a little while I can.

By the way, has anyone noticed that the Atlantic hasn't had any hurricanes yet? They usually start in late May. They don't threaten my location (too far inland), but they do add some welcome variation to the weather.

Don't miss my Windows 10 hints published yesterday.

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