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

Daily Notebook

Popular topics on this page:
How the political primaries are failing us
WPF RichTextBox.Foreground is slow
The girl who wouldn't share her GPA
An unnamed nebula in Monoceros
Stupid Windows 10 trick   overheated laptop!
Daylight saving time? The data
Directions in outer space
Making Stellarium show the Great Red Spot correctly
Why is it called a television set?
Making R display a frequency table by rows
Why are some Weka classifiers grayed out?
Are all red things fire trucks, or vice versa?
About arguing
Safe spaces and trigger warnings: A media myth?

Astrophotos:
Moon (Alphonsus)
Moon (Agrippa, Posidonius, Fracastorius)
M81, M82
NGC 2264 (Cone Nebula)
IC 443 (Jellyfish Nebula)
Lower's Nebula (Sh 2-261) and NGC 2169
Monkey Head Nebula (NGC 2174)
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

2016
March
28-31

Safe spaces and trigger warnings: a media myth?
And how Fox News reported a nonexistent event

It has become fashionable to say that today's college students are intellectual cowards — that they hide behind trigger warnings and safe spaces to avoid being challenged. This is a recurrent theme in some news media and many people's gossip.

The trouble is, I don't actually see this happening. I taught college for 31 years and never encountered it. I've just been asking colleagues, including a distinguished educational psychologist, and they haven't encountered it either.

"Trigger warnings" aren't about protection from challenging ideas anyhow. They're opportunities for people to brace themselves for gruesome scenes in literature and drama, or other things that might be upsetting due to their particular circumstances.

"Safe spaces" are intended to protect against harassment, not intellectual challenges or dissent. My problem with them is from the other direction — I think the entire campus needs to be "safe," not just designated parts of it.

In any case, no matter what you think about these things, and even though there have surely been silly overextensions of them here and there, we aren't seeing students regularly using them to avoid educational requirements. I'm not saying no one ever has. Once an idea gets out, the laziest students will use any excuse they've heard of. But that is not a new trend. I got more flimsy excuses in the 1980s than after 2000.

What turned my attention to all this is a very strange incident. A few days ago, Fox News reported something that didn't happen, and kept harping on it, to help promote the myth.

A few nights ago, Emory's campus was bedecked with chalk graffiti that said not only "Trump 2016" but also "Accept the Inevitable" and appeared to be part of an ongoing problem that Emory has been having with harassing graffiti (swastikas, etc.). It sparked a counter-protest, and one of the protesters said he received a death threat. Full details are unclear; it's a messy situation.

But Fox News latched onto something that definitely wasn't real. Fox TV reporters said repeatedly (with laughter in their voices) that Emory students were so "traumatized" (they first spelled it "traumitized") by chalk graffiti that they were offered "emergency counseling" by the Student Government Association.

That never happened. Student governments don't offer counseling in the first place; they don't have employees. Nor did any other part of Emory offer "emergency counseling," nor did students request it, as far as Snopes or I can determine.

But Fox kept airing the news story, apparently because it fit what they wanted to believe. Bad form, Fox. Issue a retraction.



Web makeover

The Covington Innovations web site has had some extensive revision, including a new software section. Please cruise around it and let me know if you find any problems.

I know that many older pages contain obsolete links; that's not what I'm looking for. But please tell me if you encounter missing pictures, malformed pages, or error messages.

Long-time readers will be glad to see the old CGI hit counters ("You are visitor number..."), which haven't worked for a while, have all been taken out.

2016
March
27

2016
March
25-26

Good Friday

I want to wish everyone a blessed Easter Triduum. Those who want to know what Easter is like for Christians might appreciate the fictionalized version in C. S. Lewis's Chronicles of Narnia: "If a willing Victim that has committed no treachery is killed in a traitor's stead, the Stone Table will crack, and even death itself would turn backwards."

An extraordinary amount of work has come my way, and I won't be blogging a lot for the next few days. (Drought or deluge; now it's a deluge!) See also the next item (below).



Arguing is not my hobby

My Facebook readers have been seeing the words "Thread pruned" a lot. That means I've cut off a branch of a conversation that I didn't have the time or inclination to carry on, usually one in which someone wanted to argue at length. My Facebook page is not an open forum for the whole world, and I am not challenging everyone to debate everything I say.

Some people are natural arguers. They reply to everything with a counter-argument, and then it becomes my job to respond. Or does it? That's one reason I prune threads. Arguing takes time, and I have other things to do.

Thinking of a counter-argument is an excellent practice — it's much better than just swallowing everything you hear — but, most of the time, critical thinking should take place inside your own head. Instead of demanding that the other person refute you, try to figure things out for yourself. Once you've understood both sides, then give a reasoned reply.

Some people have so little confidence in their own intelligence that they are afraid to try to understand both sides of a controversy. They're afraid they'll end up hopelessly confused or on the "wrong" side, so they try to shout down (rather than thoughtfully consider) the other side in any discussion. This is not helpful for the rest of us.

Arguing is not what I come to Facebook (or any other forum) for. I don't enjoy doing it. I would much rather learn and understand. I can argue when I have to — I got an award for debating in high school and have only gotten better since — but that's not how I want to spend my time. I don't learn much when I'm merely defending my own position against attackers.

Most importantly, it's not news to me that you have an opinion. Everybody has an opinion. When I mention a difficult question, I'm not taking a poll to see if you prefer Cruz or Clinton, Wendy's or Burger King, Chevy or Ford, or Elvis or the Beatles. The question is whether you can offer information and reasoning that is not already in the discussion. If so, I'll gladly listen. But if not, you don't really have anything to tell us.

Some news media cater to people like to keep everything simple, and who achieve this by ignoring facts. If you know only part of the story, and if you feel you have a license to ignore the rest, then it's easy to have very strong feelings. Little brainwork is required. But spreading inaccurate versions of the facts, for the sake of your feelings or anything else, is wrong, and it's a game I won't play.

Even worse, some people enjoy feeling angry, and they consume a steady diet of incomplete information and one-sided arguments to feed their anger. "That am not I" as Chaucer put it.

There was a time when I didn't use Facebook and seldom used online forums; I published my opinions only in this blog. I was glad to get e-mail from anyone who could actually point out mistakes or points I had overlooked. But getting jumped on by people merely because they had different opinions (based on less information) was not part of the experience. I'm inclined to go back to doing things that way.

2016
March
20-24

About that hat...

I've written about my straw hat that apparently attracts women...

...and I've joked with Melody that if the hat were separated from me, women would chase the hat, not me.

Yesterday that literally came true. In downtown Athens, the wind caught my hat and blew it nearly half a block to the edge of a busy street. An attractive undergraduate girl chased it, caught it, and presented it to me while a group of ladies who were coming up the sidewalk applauded.

2016
March
19

(Extra)

A word we need: predicend

In mathematics, statistics, and artificial intelligence, we need a word that means thing that is to be predicted. I propose to call it the predicend (pronounced PRED-i-send), from the Latin future passive participle praedicendum. Google tells me I'm not the first to think of this, although the word is rare.

Not predictor because it is the thing being predicted, not the thing doing the predicting. Not predicted because it hasn't been predicted yet.

For example, if I predict how much of a loan will be paid back, the amount paid back is the predicend. (That is not exactly the business AI problem I'm working on at the moment, but I have to keep it confidential.)

Latin future passive participles come into English ending in -nd. They denote something or someone to whom something is going to be done. Words in this class include:

ordinand 'person about to be ordained'
graduand 'person about to be given a degree'
addend 'number about to be added'
dividend 'number or quantity about to be divided (up among stockholders)'
integrand 'function about to be integrated'

Do not confuse these with present active participles, which end in -nt. They denote something or someone who is doing something, such as:

complainant 'one who is complaining'
litigant 'one who is bringing a lawsuit'
decedent 'one who is dying (or rather has died)'

They need not be nouns; they also come into English as adjectives, the Latin equivalent of words that end in -ing, such as:

prominent 'standing out'
fervent 'boiling' (i.e., intense)
adolescent 'entering adulthood'
dominant 'dominating'

and many others.

2016
March
19

Moon (Agrippa, Posidonius, Fracastorius)

Here are some more moon images, all taken at the f/10 focus of my 8-inch EdgeHD with a Canon 60Da in movie crop mode. Each is the stack of the best 75% of more than 7000 frames, processed with PIPP, AutoStakkert, RegiStax 6, and Photoshop.

From the evening of March 16, the area of the crater Agrippa:

And from the evening of March 13, the areas of the craters Posidonius and Fracastorius, in that order:

 



Are all red things fire trucks, or vice versa?

(From a Facebook conversation.)

Can you distinguish "All fire trucks are red" from "All red things are fire trucks"? If not, you might be taken in by fallacious arguments like one I saw on a web page.

The web page started out by asserting that 97% of people with terminal cancer have had root canals in their teeth, and that's supposed to prove that root canals cause cancer.

I don't think that claim was true, and because the web page was full of malware, I couldn't read any more of it. I'm not going to be able to go back to it. I think it's 100% bogus.

But even if what it said is true, it doesn't prove the point. We don't want to know how many cancer patients have had root canals. We want to know how many root canal patients got cancer. (And how many didn't.) Otherwise it's exactly like concluding that all red things are fire trucks when all you know is that all fire trucks are red.

For most people, this distinction is obvious. But some people have a lot of trouble with it. As any teacher knows, a few percent of human beings are surprisingly unaware of logical connectives. They can connect redness to fire trucks, but they can't exactly tell you which one implies which.

Others are easily bamboozled by unfamiliar or impressive terminology. They see "97%" and "cancer" and "doctor" and throw their common sense out the window.

2016
March
18

A busy time

All of a sudden I have a big software project in hand, so I'm busy. Some people don't realize that after retiring from the University of Georgia, I went into business as a software developer. That's what Covington Innovations is all about.

So today I will actually use the Daily Notebook as a notebook, to record and share a couple of tips related to the work I'm doing.



Making R display a frequency table by rows

In R, frequency tables (one-dimensional contingency tables) are normally displayed with a column for each value, with the number of occurrences of that value beneath the label.

Particularly when the labels are long, this is awkward, and you'd rather have each value on a row.

This is accomplished (strange to say) by as.matrix, as follows:



Why are some classifiers in Weka grayed out?

This is very basic, but when I was getting started with Weka, it took me a long time to get a straight answer, so I'm writing it up for posterity on the Web.

When you are going to run classifiers in Weka, you often find that some of them are not available (grayed out). In fact, you can select the grayed-out classifiers, but then the Start button is grayed out.

Obviously, the classifiers are grayed out because their requirements are not being met. But how do you find out the requirements?

What you do is hover the mouse cursor on the classifier, and look at the "CAPABILITIES" section of the window that pops up.

Apparently, in New Zealand, "capabilities" means "requirements." (No... but it does strike me as an odd choice of words.) Here we learn that MultinomialNaiveBayes requires the class (the feature you're trying to predict) to be a binary or nominal class, with missing values permitted; and it requires all the other features (the attributes) to be numeric. And that the minimum number of instances for each attribute is 1, which means there is no minimum.

2016
March
17

Lunar craters: Ptolemaeus, Alphonsus, Arzachel

Top to bottom, these are the lunar craters Ptolemaeus (Ptolemy), Alphonsus, and Arzachel. The mountain peak in the middle of Alphonsus has been suspected of volcanic activity (unlikely) or some other kind of strange behavior (perhaps outgassing from rocks).

This was taken at the f/10 Cassegrain focus of my C8 EdgeHD. It is a stack of the best 75% of more than 7,000 video frames, processed with PIPP, AutoStakkert, RegiStax 6, and Photoshop. When this many frames are stacked, the image is delightfully grainless, and the large number of frames overcomes any information loss from Canon's video compression; at the same time, because of the compression, the video files are small enough that they don't constitute a storage problem.

(Note that compressing video need not be very lossy; great compression can be achieved, in an image like this, by transmitting only the differences between successive frames.)

A further advantage of using a DSLR as an astronomical video camera is that, before switching to "movie crop" mode for the actual high-resolution image, I can use the camera's viewfinder with its full field for framing and focusing. No flip mirror or high-power finder is needed.

2016
March
16

Why is it called a television set?

Why is a TV receiver called a television set? Set of what?

Someone on Facebook asked, and I have the answer.

Television set is analogous to radio set.

Before about 1930, a radio was not necessarily something you bought as one appliance. Rather, like a very early microcomputer, it consisted of parts you could buy separately: antenna, crystal or tube detector, amplifier if any, and headphones or speaker.

A packaged combination of these, ready to use together, was called a radio set, not unlike a model train set or a set of tools.

Now you know!

2016
March
15

Giving directions in outer space

On Sunday, a young man at church asked me a question that turns out to have hidden depths. It is this: How do we describe directions in outer space? Can you tell a spaceship pilot to go north or south? Or does he use something like latitude and longitude? Or what?

A very short answer: In space, "north" means the direction in which the earth's north pole points. Imagine a line from the center of the earth through the north pole, extended indefinitely into space. That direction is (geocentric) north. So, yes, a spaceship can go north.

A better answer is that, in general, we describe the location of objects in space as latitude, longitude, and distance.

Think of a distant star. Draw a line from the center of the earth to that star. Your line comes through the surface of the earth at a particular place. (If you stood in that place, the star would be directly overhead.) And that place has a latitude and a longitude. The other thing you need to know is the distance to the star; then you've specified exactly where it is.

A minor complication is that the earth turns and carries its lines of longitude with it. (Latitude isn't affected.) We would rather have a system of longitude that doesn't rotate with the earth. To do that, we pretend that the earth isn't rotating, that it is always in a fixed position.

Latitude in the sky is called declination and is often called positive and negative for north and south. For instance, stars at declination +34 degrees pass directly over Athens, Georgia.

Longitude in the sky is called right ascension (it's a long story why) and — surprise! — is measured in hours, not degrees, taking 24 hours as a full circle. That makes it easy to figure out how long it would take the earth to rotate from one position to another.

(The strange term right ascension literally means "rising straight up" and goes back to ancient times. Longitude in the sky was discovered in a roundabout way, through people trying to calculate when stars and planets would rise if they were viewed from the equator rather than from the observer's actual latitude.)

What I've just described are called equatorial coordinates because they are based on the earth's equator. Now the fun begins.

Long ago, people realized that if you're going to pretend the earth isn't rotating, you can also pretend that its axis points in some direction other than it really does. This could be useful.

More than 2500 years ago, star gazers started using ecliptic coordinates, consisting of latitude and longitude measured with an axis that is tilted so that its equator is the plane of the solar system (specifically, the earth's orbit around the sun), and its poles are, of course, perpendicular to that imaginary equator. As you might imagine, this is very handy for tracking the orbits of the planets.

Because it doesn't have anything to do with the rotation of the earth, ecliptic longitude is measured in degrees, like latitude, not in hours.

A third system, adopted in the 20th Century, uses the plane of our galaxy (the Milky Way) for its equator, and the direction of the galactic center for its prime meridian. These are called galactic coordinates, and their great advantage is that they are not affected by the orbit or rotation of the earth. The galactic coordinates of a star remain the same forever, except for the star's actual movement, whatever it may be.

So those are three different latitude-longitude-distance coordinates.

The great advantage of latitude-longitude-distance coordinates is that you can aim a telescope at an object without knowing the distance. The latitude and longitude give you the direction as seen from earth.

The disadvantage is that if you need to know how far apart two objects are, or what direction something is moving in space, you have to do trigonometry.

If you want to visit something in a spaceship or apply the laws of physics, you need to turn latitude-longitude-distance coordinates into X-Y-Z coordinates corresponding to ordinary 3-dimensional space.

To understand X-Y-Z coordinates, imagine three perpendicular lines passing through the center of the earth. One goes through the north and south poles; one goes through the equator at longitude 0; and one goes through the equator at longitude 90.

Let these three lines be infinitely long, and measure distances on them in both directions.

Now you can use them as the axes of a graph and treat space as 3-dimensional graph paper. Everything has a distance in X, Y, and Z. This is three-dimensional space as we normally think of it.

Those are equatorial X-Y-Z coordinates. There are also ecliptic and galactic X-Y-Z coordinates. Ecliptic X-Y-Z coordinates are very important in computing the orbits of planets and in steering spacecraft. The rest of the time, we rely on earth-centered latitude-and-longitude coordinates to tell us how to aim our telescopes.

Astronomers will note that I have distorted the story in one particular way. I've been talking as if people in ancient times measured latitude and longitude on the earth first, then transferred it to the sky. That is actually very hard to do; although the ancient Greeks knew the earth is round, they didn't know that it rotates, and they didn't have much ability to measure positions on it.

So, instead, they thought in terms of the celestial sphere. In ancient times, people imagined that the stars were on a huge globe surrounding the earth which rotated around the earth once per day. It's easy to imagine drawing lines of latitude and longitude on the celestial sphere, and you quickly realize that they need not have the same axis as the earth. That is how ecliptic coordinates came to be invented many centuries before Christ.

Today we realize the celestial sphere is not real. But it remains a convenient fiction; I use it nearly every time I calculate positions.



How to make Stellarium show Jupiter's Great Red Spot correctly

[Updated.]

I use Stellarium as my all-purpose sky map. When I started observing Jupiter this season, it didn't take me long to find out that although Stellarium shows the Great Red Spot, it doesn't necessarily show it in the correct position.

But yours may be working just fine, if it's the latest version and doesn't have any files carried over from earlier versions.

Anyhow, here's what to check.

[Added:] Before anything else, go to the Sky and Viewing Options menu and check "Simulate light speed." It takes light about 45 minutes to get to us from Jupiter. If you don't account for that, you'll be seeing everything — satellite positions as well as the Great Red Spot — slightly wrong.

First, regardless of which fix you're going to use, take a look at the two files:

C:\Program Files\Stellarium\data\ssystem.ini

%appdata%\Stellarium\data\ssystem.ini

They may be identical. If not, and if you have not customized this file, copy the newer one from Program Files into %appdata%.

Next, open ssystem.ini with Notepad++ (not Notepad, because it has UNIX end-of-line marks).

Look for the section that begins [jupiter], and within that section, the line that begins with rot_rotation_offset.

Now you have a choice.

If you specify rot_rotation_offset=-1 you will activate Stellarium's built-in algorithm for locating the Great Red Spot. It is quite good, though not perfect. The Great Red Spot drifts, and you should update Stellarium regularly (and copy the latest ssystem.ini) if this is the algorithm you're using. The position of the Great Red Spot gets updated when you do.

Otherwise, you can adjust the rotation of the Jupiter image yourself. I did, and with Stellarium 0.14.2, I am getting good results with:

rot_rotation_offset=35

That may work for you. Otherwise, you'll need at least one recent picture or observation of Jupiter with the Great Red Spot visible and an accurately known date and time (to within perhaps 5 minutes), or a date and time from a table of Great Red Spot transits.

Specify a value for rot_rotation_offset in degrees, save the file, launch Stellarium, set the date and time to the picture you're trying to match, find Jupiter, and zoom in.

If the Great Red Spot isn't on the disk, try changing rot_rotation_offset in increments of 60 degrees until it is. Then make smaller changes until it's in the right position. Every time, you'll need to save the file, close and reopen Stellarium, set the date and time, and zoom in on Jupiter.

Then remember to go back and do the adjustment again every few months, or at least once every season. The Great Red Spot is drifting about 14 degrees in longitude per year, but the drift is subject to a lot of irregularity.

Technical note: For Jupiter, rot_rotation_offset is not a simple offset in longitude, as it would be for any other planet. Instead, it is the input to a mathematical model of the observed movement of the Great Red Spot.

If you specify -1, Stellarium computes the position of the Great Red Spot from scratch, ignoring the Stellarium's rotation period for Jupiter (which is System III) and using a model of the rotation of the Spot itself (which is not 100% predictable, but they do rather well).

If you specify a positive number, the value you give will still be adjusted for estimated long-term drift, which may puzzle you as you try to adjust it yourself. Indications are that in version 0.15, Stellarium will no longer try to do this, and the offset will simply remain fixed at the value you specify; your adjustments in subsequent months will then need to match the actual drift of the Great Red Spot.

I want to thank members of Cloudy Nights and Stellarium discussion for useful information about this.

2016
March
14

Today on "Dad the Electrician"...

What I posted on Facebook yesterday:

Started to install an under-cabinet LED light strip in our breakfast room.

Noticed that the outlet was not holding the plugs tightly. Replaced outlet.

Wanted air conditioning. Turned air conditioner on. Noticed that it wasn't running.

Went outside. Air conditioner compressor wasn't running because a huge rubber cable conduit had fallen to block the rotation of the fan. Observed that an enormous zip-tie, placed there a few years ago by the repairman, had broken.

Cut power to compressor. Using coat hanger and grabbing tool, pulled that conduit back to where it should be. Verified that the fan blades had only nicked the outside of the conduit, not any wires. Reinstalled it in place using 4 big zip-ties instead of one enormous one.

Came back inside. Air conditioning works. Finished up securing the cord for the under-cabinet light.

Noticed that the under-cabinet light gets in my eyes when I'm sitting at the breakfast table. Relocated it 3 inches west of original location.

Now taking a break.

I pointed out to Melody that all these repairs were made possible by my long history of buying things at hardware stores that looked useful even in the absence of an immediate need.



One more thought about Daylight Saving Time

I've noticed that people's feelings about Daylight Saving Time seem to depend on the extent to which they appreciate the variation of sunrise and sunset as part of nature.

For example, do you expect noon to be halfway between sunrise and sunset? For those of us who expect clocks to describe nature, that's what noon is; it could hardly be anything else. (When the day gets longer, it gets longer in both the morning and the evening.) But for those who think only in terms of human activities, sunrise should always be at 7:30 a.m., and the inability of Daylight Saving Time to make it so is frustrating.



Gluten-free...

I seem to have made up a joke that will continue reverberating around the linguistics profession for a long time.

Is the opposite of an agglutinative language (one with lots of single-purpose prefixes and suffixes) a gluten-free language?

2016
March
13

Daylight saving time? The data

Daylight Saving Time is a genuine burden in not only inconvenience but also health and safety, and we should debate whether to continue it.

People on both sides of the debate may appreciate some actual data about its real effect. So today I'll give you a table of sunrise and sunset times, with and without DST, followed by discussion. The biggest conclusions that I draw are:

  • Daylight Saving Time can be understood as an attempt to hold the time of sunrise constant, so that all the extra length of the day will be at the end.
  • Nonetheless, over two thirds of the extra sunlight after dinner is purely natural, not caused by resetting the clocks. It would still be there if we didn't have DST.
  • Year-round DST would cause children to have to go to school in the dark, as long as we keep the tradition of starting the school day at 8:00 or 8:30 a.m.
  • Georgia and Florida are in the "wrong" time zone, but they may need to stay there because of the importance of Atlanta as a transportation hub.

More about some of these points shortly. First, some sunrise and sunset times:

Dec. 21
Shortest
day
Mar. 21
Average
day
June 21
Longest
day
 
The way we're doing it now (DST in summer, not winter)
Atlanta Sunrise 7:39 a.m. 7:39 a.m. 6:27 a.m.
Sunset 5:33 p.m. 7:50 p.m. 8:51 p.m.
Los Angeles Sunrise 6:55 a.m. 6:55 a.m. 5:42 a.m.
Sunset 4:48 p.m. 7:06 p.m. 8:08 p.m.
Minneapolis Sunrise 7:48 a.m. 7:13 a.m. 5:27 a.m.
Sunset 4:35 p.m. 7:28 p.m. 9:03 p.m.
 
Without Daylight Saving Time
Atlanta Sunrise 7:39 a.m. 6:39 a.m. 5:27 a.m.
Sunset 5:33 p.m. 6:50 p.m. 7:51 p.m.
Los Angeles Sunrise 6:55 a.m. 5:55 a.m. 4:42 a.m.
Sunset 4:48 p.m. 6:06 p.m. 7:08 p.m.
Minneapolis Sunrise 7:48 a.m. 6:13 a.m. 4:27 a.m.
Sunset 4:35 p.m. 6:28 p.m. 8:03 p.m.
 
With year-round Daylight Saving Time
Atlanta Sunrise 8:39 a.m. 7:39 a.m. 6:27 a.m.
Sunset 6:33 p.m. 7:50 p.m. 8:51 p.m.
Los Angeles Sunrise 7:55 a.m. 6:55 a.m. 5:42 a.m.
Sunset 5:48 p.m. 7:06 p.m. 8:08 p.m.
Minneapolis Sunrise 8:48 a.m. 7:13 a.m. 5:27 a.m.
Sunset 5:35 p.m. 7:28 p.m. 9:03 p.m.

Now, some observations.

Atlanta and Los Angeles are at the same latitude. So why does everything happen 45 minutes earlier (local time) in Los Angeles? Because Atlanta is in the wrong time zone. Georgia and most of Florida are at a longitude that should be in the Central Time Zone, but they've opted to be in the Eastern Time zone. That is equivalent to having about three quarters of an hour of DST year-round, plus another hour on top of it in the summer.

Arizona has the same problem and has opted not to have DST. As a result, Arizona matches New Mexico on its east in the winter and California on its west in the summer. It's tempting to say Georgia should make the same compromise. However, Atlanta is such an important transportation hub that Georgia probably shouldn't do anything that people would have trouble keeping track of.

Turn your attention now to Minneapolis and Los Angeles, which are more or less centered in their time zones, but one is much farther north. In Minneapolis, the time of sunset varies over a four-and-a-half-hour range, only one hour of which is due to DST. That's right — even if there were no DST, sunset would be three and a half hours later in the summer. Daylight Saving Time doesn't make as much difference as most people probably think.

In Los Angeles (and Atlanta) farther south, the variation is only three and a quarter hours, one hour of which is due to DST. So without DST, we would still have appreciably later sunsets in the summer, though the difference is less dramatic.

If you like late sunsets, what about year-round DST? The problem then is that winter sunrise would come well after 8 a.m. and children would have to go to school in the dark. Of course, one might argue that the school day (and work day) are starting too early.

I would advocate dropping DST nationwide. Besides of the health and safety burden of which previous generations were unaware, we also have the inconvenience of resetting what can easily be dozens of clocks in every household. (I have to go reset three cameras and two voice recorders right now, as well as clocks in the proper sense of the word.) These are not like Grandpa's clock that had to be set every week; they are quartz clocks that would easily keep time for a year if left alone.

Most of all, I want to reassure people that if we drop DST, the days will still be longer in the summer. Over two thirds of the extra sunlight after dinner is due to nature, not clock tricks.

2016
March
12

Stupid Windows 10 trick — overheated laptop

[Revised.]

Windows 10 can actually damage your laptop. For details, read on.

Today's Bad Engineering Award goes to Microsoft for the following misfeature of Windows 10. Quite a few earlier versions risk the same problem, but the Windows 10 makes it worse.

If a laptop, in standby (sleep) mode, is placed in a closed bag, it may overheat and be damaged because it will wake up in the middle of the night and install updates, even though it has no ventilation.

This happened to my daughter, Catherine Barrett. Fortunately, her laptop detected the overheating and shut down. But I don't know if all laptops do, and I don't know to what extent the life of components may have been reduced by this misbehavior. After all, by the time the CPU detects overheating, things are hotter than they should be!

All earlier versions of Windows gave you the option of just being notified of updates and not installing them until you want to. Not any more.

In Windows 10, you can still suppress automatic updating by disabling your Internet connection or by telling your computer that the connection is metered (pay-by-the-megabyte). Details here. Most people would probably never guess that this is how it's done.

Microsoft, please give us more choice about these things!

To keep automatic updates from waking up a sleeping computer, you can apparently use some well-hidden settings pertaining to "wake timers." (See this Microsoft page.) Under Settings, System, Power, Power & sleep, Additional power settings, Change plan settings (for the plan that is selected), Change advanced power settings, you will find Sleep, Allow wake timers. Turn it off.

It is also a good idea to set the computer to hibernate rather than sleep when you power it off. That is under Settings, System, Power & sleep, Additional power settings, Change what the power buttons do. This keeps it truly powered off, rather than running at a low level. However, I am not sure whether this by itself will prevent Windows Update from waking it up. (If anyone can tell me, please do.)

Ideally, of course, a computer that is placed in a bag for storage should be completely powered down, not just asleep or hibernating.

2016
March
11

An unnamed nebula in Monoceros

Update: For better pictures see April 26.

I want to call attention to the reddish patch indicated by the arrow in the picture above. It is a nebula that has no catalog designation, only the positional designation GAL 201.6+01.6 (SIMBAD) or G201.6+1.6 (IAU style), based on its galactic coordinates. I wrote about it briefly in January.

Not totally unknown to science, this nebula has been studied as a radio source (all ionized hydrogen nebulae emit radio waves) and is estimated to be 4.4 kiloparsecs away, where I've plotted a big red dot on NASA's artist's conception of our galaxy. It's apparently between the Orion Spur and the Perseus Arm, or maybe in one or the other of them, depending on how uncertain the distance is.

What I don't know is whether the nebula can be seen visually with amateur telescopes. Since it eluded Sir William Herschel and all the other 18th- and 19th-century observers whose observations were combined to make the NGC and IC catalogues, I have my doubts, but if you have a dark location and a good-sized telescope, you might try. It's a quarter degree west and a little south of the 6th-magnitude star SAO 95914, which is the bright star at the top center of the picture. J2000.0 coordinates of the nebula are 06 36 36.3, +10 46.

2016
March
9-10

Lower's Nebula (Sh 2-261)

Here is a better picture of Lower's Nebula, whose story I told back in January, along with the "37 cluster" (NGC 2169, which looks like the number 37) south of it.

Stack of 35 1-minute exposures, Canon 60Da, Canon 300-mm f/4 lens, on Celestron AVX mount, no guiding corrections.



NGC 2174

Here is NGC 2174, the Monkey Head Nebula (not looking particularly pithecine in this picture). This is the nebula that photobombed my picture of M35 back in November 2014. At that time I labeled it NGC 2175, which is the star cluster inside the nebula.

Stack of 40 1-minute exposures, Canon 60Da, Canon 300-mm f/4 lens, on Celestron AVX mount, no guiding corrections.

2016
March
7-8

IC 443 (Jellyfish Nebula)

Continuing our tour of emission nebulae (H II regions) in the Orion spur of our galaxy, in the direction of the constellations Orion, Monoceros, and Gemini, here is IC 443, the Jellyfish Nebula, so called because it looks like a side view of a jellyfish. Photographed under a dark country sky, it would look even more like a jellyfish, with streamers hanging down (i.e., to the right in this picture), but this picture was taken in town and misses the faintest parts of the nebula. One day I'll do it again at Deerlick.

Stack of twenty-five 1-minute exposures, Canon 60Da and 300-mm f/4 lens.



Short notes

I'm gratified that my essay about the girl who wouldn't share her GPA has been well received by both liberals and conservatives and is even starting to circulate in academia.

At Covington Innovations, business is rumbling, if not actually booming. Some major consulting projects have been delayed; other possible projects are popping up; and (as I sometimes say) "the light on top of the taxi is on," i.e., I'm open to starting more projects.

2016
March
6

Fake tax document of the day
("We tricked you! Now trust us!")

I don't know whether to describe a certain kind of advertiser as dishonest or just afflicted by a strange blind spot about right, wrong, and how people will react to things.

Look what came in the mail the other day:

Although it closely imitates the appearance of the various IRS W-2 and 1099 forms that we're receiving this time of year, this is not a tax document. It's an offer to lend me money. The fine print says as much, once you read it.

My question is, what do they think they're doing?

Obviously, tricking me into reading their offer by making it look like something else.

But am I supposed to like this, once I see what they've done?

Has it ever entered their little heads that I might be put off by it?



Kar(ma) wash

Word to the wise: If you suddenly, aggressively cut ahead of me after we both enter the car wash, you will get the cheap wash that I paid for, and I will get the possibly more expensive wash that you paid for.

This happened to me the other day, except that actually, both of us had paid for the cheapest wash, so there was no damage done. Maybe next time...

2016
March
5

Liberals, conservatives, and the girl who wouldn't share her GPA

[Revised after online discussion.]

There's a subtly misleading story spreading on the Internet. It's about the difference between "liberals" and "conservatives."

A college girl of the "liberal" persuasion advocates sharing resources with people in need, until her "conservative" father asks if she will share her high grade-point average (GPA) with a less industrious student. "No, of course not," she says, "I worked hard for that GPA." Her father replies, "All right, then, welcome to the Republican Party."

What's wrong with this picture?

At first sight, nothing. Obviously, GPAs shouldn't be shared. Obviously, it would be unjust to lower one student's grades and raise another student's grades because the latter one is "less fortunate."

In fact, the story does make one point well: forced leveling destroys incentive. This is now well known to economists. (See also this.)

But wait a minute. The story has some flaws.

Flaw #1 in the story is that GPAs don't work like money. A GPA isn't a medium of exchange or a tradeable resource. It's a measurement, and you can no more share it than you can share your shoe size. If you try, all you'll do is make it an inaccurate measurement. That is why the notion of sharing a GPA is so repugnant.

So the story makes it seem that money is only a reward for achievement, a status symbol. It overlooks the fact that money is the means of obtaining necessities, a resource that we hold and use for the good of others.

Flaw #2 is in the definition of "liberal." It is an instance of a common mistake, assuming the "other side" in a controversy is the exact opposite of everything you stand for, as if it were the opposing football team, obligated to run the ball in the opposite direction. In reality, on the stage of real-world politics, opposite sides usually have quite a lot in common; they differ about details.

Does anyone, other than a few wild-eyed left-wing extremists, actually advocate taking income and incentive away from hard-working people and giving it to lazy people in order to force equal outcomes? No. That kind of naive Marxism is dead, or at least isn't called "liberal" on today's political stage.

Today's liberals and conservatives equally recognize the advantages of the free market. Liberals are more sensitive to market failures — situations where the free market doesn't do its job, typically because people lack freedom or information or are prejudiced. They are also more sensitive to externalities, which are costs imposed on people against their will, such as environmental damage from pollution. And they are more optimistic about the ability of government to fix these problems.

Liberals are also more concerned about income inequality. But today the focus is on raising up the poor, not knocking down the rich, and on the question whether rising income inequality is pulling down the productivity of the whole economy. A little more about that here.

(I am, by the way, rather toward the conservative side, but let's be fair in how we characterize those with whom we disagree.)

So the real effect of the story is to stir up contempt for "liberals" by saying they're something they aren't. And knowingly spreading a false characterization is wrong.

Flaw #3 has to do with the lesson we're supposed to take away. It's actually a bundle of things. That hard work should be rewarded? Surely so. That leveling destroys incentive? Yes, indeed. That helping the less fortunate is a bad thing? Wait a minute, not so fast!

The left and right agree that we need a "social safety net" (a term introduced by Reagan, who was no leftist) to intercept people who are falling into poverty and break their fall, so they can return to productivity if possible, and if not, at least not burden their immediate families and neighbors. This is good for the whole economy, not just the people who receive benefits. Its purpose is not to play Robin Hood, to level incomes, but rather to rescue people from real calamities, just the way insurance does.

Lots of people imagine the social safety net is full of lazy freeloaders, and yes, there are a few. But a few examples do not prove anything about large numbers. And honest people will not use imagination as a substitute for facts.

There is a serious problem with the cost of entitlement programs — if the costs of a program are growing faster than the source of the money that should fund it, it's going to run out of money, and no amount of good intentions will change that. But this is not because of lazy freeloaders. It's because of bad design.



Downing on Piketty

[Revised; I thank a correspondent for pointing out an unclear passage.]

I want to share with you a bit more from Freedom, Opportunity, and Security, by my good friend Doug Downing. Unlike me, he's an economist.

Economists and politicians are concerned about growing income inequality — a widening gap between the rich and the poor. A French economist, Thomas Piketty, thinks he has figured out why inequality is growing and what to do about it. The discussion breaks down into three issues:

(1) Does Piketty's economic analysis make sense?

(2) Do the facts support it?

(3) Is inequality a problem anyhow?

Starting with (3), beware of thinking that there's a fixed amount of wealth to go around, so the only way to make the poor richer is to make the rich poorer. That's not how it works. By the world's standards, I'm rich; I have a much bigger house than the average resident of Planet Earth. But where was my house 50 years ago? Did I get it by taking it away from someone else? No; it was built in 1974. People labored in 1974 to create it. If they had not done so, it wouldn't be someone else's house, it would be nonexistent. Other things of value work the same way.

Beware also of wishing for poverty for equality's sake.

If these are the incomes [of 5 people, the entire population] in society A:

90, 100, 100, 100, 110

and these are the incomes in society B:

500, 600, 1000, 1500, 2000

it is clear that society B has much more inequality, but society A has much deeper poverty. ... It seems perverse to prefer a society where everyone is poor (as in A) over a society in which some people have gotten very rich while others are somewhat rich (as in B). ...

If a country used to have all of its people being extremely poor, and now it has half of its people much better off while the other half is still extremely poor, do you lament the fact that the country is now much more unequal than it used to be, or do you celebrate that many people are now less poor (and then wonder about how to help the rest of those people become less poor)? [p.37]

One can also ask if the Titanic would have been better if all 2200 people had died (equality) than in the actual situation where 700 survived:

Hopefully you realize that the problem was that 1500 people died, not that there was inequality... Imagine someone writing, "Inequality on the Titanic increased rapidly during the period when some people left the sinking ship to board floating lifeboats."

Piketty actually does write, "Chinese inequality increased very rapidly following the liberalization of the economy in the 1980s [...]." There are two problems with this statement. Is it really true that China under Mao had equality? It is true that just about everybody was desperately poor, but one person [the dictator] had unlimited power. A society with one person with everything and everyone else with nothing is an example of the maximum possible inequality. For the sake of the argument, leave aside this point... In China before Deng Xiaopeng, just about everyone was poor. Under Deng and his successors, literally hundreds of millions of people have become fairly rich. ... There are still hundreds of millions that are poor, but does it really make sense to focus on the growing inequality as if that was the problem, rather than focusing on the fact that so many people are no longer poor? [p. 38]

Now to (1) and (2). Piketty's basic claim is that inequality has to increase when the rate of return on capital exceeds the growth rate of the economy. This seems to assume that the rich always invest as much as they can. But in fact what do rich people do with some of the return on capital? Spend it! Doug calls this the "what's the point principle" — why accumulate wealth if you can't enjoy it? As people get richer, they don't just keep reinvesting. And even if they do, even if all they want to do is get richer, under capitalism they can only do it by making the whole economy more productive — which is good for the poor and the working class. Piketty's analysis would work better under feudalism.

In place of Piketty's concern about inequality, Doug cites John Rawls' concern about the poorest of the poor. If an economic change improves the lot of the worst-off members of society, then it's a good thing, even if it improves the lot of someone else even more.

For more criticism of Piketty, see Wikipedia, which is frequently updated.

2016
March
4

Some other galaxies for a change

In honor of Exelaunei Day (and the Exelaunei Day greeting I just received from one of my teachers), I want to march forth out of the region of the galaxy I've been photographing, and show you something much farther away.

Here you see the galaxies M81 (center), M82 (above it), NGC 3077 (left), NGC 2976 (lower right), and, halfway between 3077 and 2976, the faint, distant galaxy UGC 5302 (15th magnitude). Except for UGC 5302, these are nearby galaxies, about 8 to 10 million light-years from us in the direction of the constellation Ursa Major. At least the first three of these are an interacting group; relatively close to each other in space, they have pulled gas away from each other. M82 has an especially irregular shape.

Stack of forty 1-minute exposures with a Canon 60Da and 300-mm f/4 lens, taken in my driveway in brightly-lit Athens, Georgia.

2016
March
2-3

WPF RichTextBox.Foreground is slow
(One reason output to a RichTextBox can be slow)

In a C# program I was working on today, output to the screen was inexplicably slow. Output was being written into a WPF RichTextBox object, and not all of it was the same color of type. The slowness seemed to be worst in windows with lots of color changes.

To make a long story short, the culprit turned out to be assignment statements like this:

box.Foreground = Brushes.Red;

For unknown reasons, such an operation takes as much as 50 milliseconds! Maybe WPF requires the GUI to handle all pending events before changing the foreground color of a RichTextBox. That's just a guess. I cannot confirm it from the published source code.

To write into a RichTextBox quickly, the proper method is to create a Run object (a run of text, with a specified color) and append it to the current Paragraph object of the current Document object. More complicated, but much faster!

The other way to speed up RichTextBox output somewhat is to set IsUndoEnabled to false.

2016
March
1

How the presidential primary elections are failing us and how to fix them

As I write this, we are waiting for Super Tuesday primary returns. National polls indicate that the Republican vote is likely to be about 30% Trump, 20% Rubio, 20% Cruz, and the rest split among other candidates.

The problem is, Cruz and Rubio are more capable of winning the general election. But they are so similar to each other that they've split the vote, leaving Trump with the plurality even though everyone outside his 30% finds him completely unacceptable.

What's wrong? Simple. Elections are designed to choose between candidates who are different from each other. But in a party primary, it is normal and desirable for candidates to be similar to each other — they're in the same party!

The current system seems designed to divide parties rather than unify them by making the party's internal process work too much like the national election.

(I don't think the Democrats are being served too well either. Their two leading candidates are too far apart — Sanders is too far to the left to win the general election, but Clinton, the moderate, may be about to face criminal charges for mishandling secret messages. Nobody politically close to Clinton is still in the running.)

Here's one way to fix it. Instead of having voters choose one candidate in the primary, have them rate each candidate as "acceptable" or "unacceptable." The one with the most "acceptable" votes wins the party's nomination.

Other systems of preferential balloting could be used. What they all have in common is allowing people to express support for more than one candidate. That way, a candidate whose following is narrowly exclusivist (Trump) would not win over a candidate whose followers also like several other candidates (Cruz, Rubio, Kasich, Carson).


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