The other day, I was attempting to sign up for some service or other using a form that demanded a ’security question’ and answer. Insecurity question would be a better term, actually, particularly as this thing gave a choice of only three questions.

The ’security question’ is used to help you retrieve or reset your password if (when) you forget it. It’s a bad idea all around, really, but if you have to do this, the least bad way is to allow the user to specify the question and the answer. When the user is trying to retrieve a forgotten password, you then display the ‘question’ string, and check to see whether what you get back is the ‘answer’ string.

If you think about it for about a millisecond, you’ll note that this is nothing but another password for the user’s account. Let’s say that my username is ‘tino’ and my password is ‘p455w0rd123′; the logic of the login is ‘This person knows that the password ‘p455w0rd123′ is associated with the username ‘tino’, so this must be Tino. He can do what he wants.’

There are conceptual weaknesses in this — primarily that most people use the same password (which they never change) for everything — but on the whole it works well enough.

The ’security question’ allows for the same kind of user identification, but it does it in an intentionally broken way. The password ‘p455w0rd123′ is arbitrary; that is, it has nothing to do with anything except that it’s easier for a human to remember than is ‘f88a2e6abe98bbef81ddb3f86c410f93′.

The ’security question’ is a password, but one that’s inherently and deliberately not arbitrary. The common canned questions ask for:

1. Your mother’s maiden name
2. The street you lived on as a child
3. The name of your childhood pet
4. Your favorite sports team

These are spectacularly bad. #1 can be defeated with a simple public records search; #2 and #3 are 100% vulnerable to anyone who knew you well as a child (or anyone who can ask questions of any of those people). #4 is particularly ridiculous, given that people trumpet their sports affiliations with hats, T-shirts, license-plate frames, etc., etc.; and in any case it’s extremely likely that your favorite sports team is one that plays in the city in which you live, or in the city where you grew up, or for your alma mater.

Even some of the ‘good’ questions proposed at GoodSecurityQuestions.com are ridiculous. These, for instance, are all useless against anyone who knows you well enough:

• What was your childhood nickname?
• What is the name of your favorite childhood friend?
• What street did you live on in third grade?
• What school did you attend for sixth grade?
• What is the first name of the boy or girl that you first kissed?
• What was the last name of your third grade teacher?
• What was the name of your elementary / primary school?
• In what city or town was your first job?

Most of the rest of them are similarly bad, vulnerable to public-records searches or simple guessing. The whole thing is a bad idea, as has been scientifically proven.

But what did I find that was even more ridiculous? I found a security question system that insisted on a minimum length for your answer, confusing the minimum-length requirements for random passwords with the specific piece of information that’s required by ’security’ questions. If you grew up on a street with a name less than six letters long, or if your mother’s maiden name was Smith, or if your childhood pet was Fido, or if you were born in Miami: you have to enter some longer variation (’Elm Street’, ‘Mary Smith’, ‘Fido The Cockroach Slayer’, ‘Miami, Florida’) and then remember how you’d padded out the answer in the future when you needed to supply the answer.

I didn’t bother to complete the registration.

While complaining about the Kindle’s typography, I tried to find an example of another typeface that I knew I’d seen on the thing at some point. After failing, I figured that this was my imagination.

Almost as soon as I posted my Kindle 2 review, though, I came across the Amazon sample of Jim Kunstler’s post-apocalypse fantasy World Made By Hand. I’m no typographer, but it looks like some Century Oldstyle variant:

The text reflows when you change type sizes, too, so it’s not like they’ve cheated by making the book up out of images.

So: there is a decent typeface on the Kindle, and the thing can already display books using it. This just doesn’t seem to be documented anywhere that I’ve been able to find, and you can’t tell the Kindle to prefer this instead of the bad implementation of Caecilia.

This is actually worse, when you think about it. The hard work has been done; all that’s needed is a switch for the user to pick a preference. That this switch does not appear to be there does not reflect well on Amazon.

Addendum to the addendum: It also appears that it’s no longer possible to choose between justified text and left-aligned text, as you could with the original Kindle by hitting ‘J’ while the type size picker was displayed. You get justified text (which generally results in the distracting spacing seen in both images above), or you get nothing.

The new Kindle in a nutshell:

• It’s the same weight as the old Kindle, but it feels heavier because it’s denser
• It doesn’t come with a case; the old case made the old Kindle much easier to hold
• I’m not sure yet whether the new next/previous page buttons are better.
• The five-way controller is better than the old wheel
• The screen is much better, thanks to an improved controller; but there are still problems with the typography

The important points are the first and last ones.

First, weight. Or, rather, density. The old Kindle wasn’t all that dense, for a modern gadget. It also had a lot of strange angles to it, which were much criticized but which almost guaranteed relatively slow deceleration should you drop the thing.

The new Kindle is denser, and it has no strange angles. The front side has a very slight bit of a dome shape to it, but that’s all. The back is also made of metal, which will hold on to more deformation — i.e. it’ll dent and warp if you drop it wrong, while the old plastic case was more likely to snap back to its original shape.

The density is important, too. Go climb up on the roof of your house, and take with you two spheres of equal weight. One of these spheres should be made of styrofoam, and the other should be made of aluminum. Aluminum, for a metal, isn’t very dense — but it’s a whole lot denser than styrofoam. Your aluminum sphere should be about the size of a baseball; I haven’t done the calculations, but I think the styrofoam sphere will be at least 36 inches in diameter. When you get up on the roof, drop both spheres off over a hard surface. I’ll wait while you find the spheres and climb the ladder.

What happened to your spheres? The aluminum one now has some gouges in it (some, not one, because it bounced), and if you have access to precise measuring equipment, you’ll notice that it’s no longer even a sphere; the whole thing deformed when it hit the ground, and didn’t go back to its original shape.

The styrofoam sphere, on the other hand, is unscathed. Its surface area is so large relative to its weight that it got significant support from the air on the way down. In fact, if it’s windy where you are, you probably had to jump off the roof and go running down the street after it. And, because of the same low density, styrofoam isn’t going to bounce — so there’s only going to be one impact to worry about.

So I think that the Kindle’s increasing density and use of metal for a back plate is a step backward. The new Kindle certainly looks better, and sleeker, than the old one, but I do not think that its construction is as practical. Marketing here has triumphed over engineering; but given that most of the criticisms of the old Kindle had to do with the thing’s appearance rather than its capabilities, maybe this was unavoidable.

Second, the screen. I think that this is exactly the same screen as the old Kindle, but with a better controller. This would be like keeping your old monitor, but buying a new video card for your computer. The new controller allows the Kindle to change only certain regions of the screen much more quickly than the old one did, and it allows for greater bit depth.

The better ability to change parts of the screen — something that happens constantly on your computer’s screen, and which you take entirely for granted — allows for a more normal interface, rather than the strange LCD-mirror arrangement for selecting items that the old Kindle used.

The greater bit depth is a huge improvement. Arguably, the only serious failure of the old Kindle was that its typography was lousy (about which more below); most of this was due to the screen’s 2-bit depth.

Bit Depth Explained

Computers think in terms of bits; a bit can be equal to either 1 or 0. Those are the only values.

For a computer to count higher than 2, it needs to use more than one bit. If you take two bits and consider their value together, you can have either 00 or 01 or 10 or 11: that’s four values. If each of those values represents a color, you can have four colors. If these ‘colors’ are actually shades of gray, you can have:

Black, white, and two in between.

Every time you add a bit, you double the number of possible values. Three bits can have twice the possible values of two bits: you have all four of the two-bit values above combined with the two possible values of the third bit.

So where your four possible two bit values are:

00 01 10 11

with three bits you have:

000 001 010 011 100 101 110 111

Add another bit, and you can have sixteen distinct values:

0000 0001 0010 0011 0100 0101 0110 0111 1000 1001 1010 1011 1100 1101 1110 1111

And so on. If you represent these as grays, you get:

That’s black, fourteen shades of gray, and white at the end.

Because of the way the eye and brain work, though, you can cleverly combine those sixteen shades of gray in such a way as to fool you into thinking that you’re seeing any shade of gray, not just the sixteen that are actually present. This strip contains nothing but the same sixteen shades of gray seen above:

There’s a bit of banding visible, but it still looks pretty smooth. Here’s a closeup of what’s actually going on there, near the middle of the strip:

All of that is well and good, but why does this matter for displaying text? Text is just black and white, right, so you shouldn’t need more than a one-bit display.

The problem is that all current displays — all the ones outside of labs, anyway — are too low-resolution for this to work well: so text on screens is anti-aliased to make it look smoother.

Here’s an example. This is a 36-point Times capital ‘T’, blown up 800%, as it would be represented on a one-bit display that can only show black and white:

You will notice that it is quite jagged.

Here is the same T, as it would be displayed on a 4-bit display that can show sixteen colors:

Looks kind of like a 3D comic without the glasses, but when it’s displayed at normal size it looks smooth:

Any digital image — like from your digital camera — is composed of individual pixels; everything is made up of little squares, each of which can only be a single color. When you take a picture of a capital A, though, for instance, the letter doesn’t come out looking like a ziggurat because the camera doesn’t resolve the boundary between the slanty lines of the A and its background perfectly; if you have a red A on a white background, you get a line of pinkish pixels between the two, and if you look at the image at normal resolutions, it looks like a sharp diagonal line.

Diagonal and curved lines in images generated by computers used to be jagged, but as computer power has decreased in price it has become common to have the computer anti-alias such things, adding intermediate fuzz to the edges in such a way that the screen displays a level of detail — as perceived by your brain, anyway — that’s greater than is, strictly speaking, possible given the screen’s absolute resolution.

This kind of thing is particularly important when you are displaying text in small sizes on the screen. The top line here is not anti-aliased, and the bottom line is what you could display with anti-aliasing on a screen capable of displaying sixteen colors:

The second line is much easier to read, despite actually being blurry. This is particularly important when you’re reading a lot of text.

That said, the improvement in the typography isn’t all that it could be. Here’s a comparison of the old and the new Kindle; the old one is on the left.

Clicking on either of those will pop up a bigger version.

The most immediate difference in these pictures is that the old Kindle has a lot more ’stuck’ globules, the little black specks in the background. This is not as obvious in real life, though, and in any case it may simply represent differences in quality control between the two screens.

The anti-aliasing is indeed much better on the new Kindle, but it’s hindered by the continued use of the Caecilia typeface.

While Caecilia is an attractive enough typeface, it is not a great face for body copy. When the original Kindle came out, I hypothesized that it had been chosen because of its slab serifs, which would work relatively well on a screen that’s only capable of crude anti-aliasing. Here’s what Caecilia really looks like:

Those square serifs make it easier to display accurately on a two-bit screen. With a four-bit screen, that’s not such a big consideration, and Amazon would do well to use a better typeface.

The particularly baffling thing is that the Kindle does have other typefaces in it, Neue Helvetica at least; there’s no compelling reason not to have a few more and to allow the user to make the decision about what typeface to use.

Other Things

The most baffling thing about the Kindle is that it continues to not be able to read HTML documents internally, despite obviously having an HTML renderer on board as part of the web browser. This means that for creating your own documents, you are limited to either:

1. Plain text;
2. Screwing around with all kinds of wacky opaque ‘e-book maker’ programs;
3. E-mailing your document to Amazon, where they apply #2 for you, with even more opacity and delay.

This is entirely nuts. Most of the complaints about the Kindle have been pointless; they come from people who haven’t used a Kindle and who don’t understand what it’s for, and who therefore complain about problems that are either 1) totally immaterial or 2) actually features.

The complaints about the page-turning buttons on the original Kindle are a good example of this. Yes, the buttons were easy to hit accidentally. But keep in mind that when actually using the Kindle for its intended purpose, the user will need to press these things thousands of times. Making them easy to press is a feature, not a bug.

I’ve seen some complaints about the new Kindle related to the five-way controller (i.e. joystick) and the menu system, to the effect that this is too difficult and/or complicated to use. What I don’t think these people realize is that you don’t use these things that much. 99% of the user’s interaction with the Kindle involves pushing the ‘next page’ button. The device is optimized for this at the expense, where necessary, of the more complex navigation.

I’ve also seen people vowing — swearing that they will not buy a new Kindle because it does not have the SD card slot that was present on the old one. This just, they say, makes the thing an entire non-starter for them. I’d prefer to have an SD card slot myself (I’m a big fan of not combining devices and data storage any more than is necessary), but it’s not that big a deal. If you’re transferring content every day, you’re not reading the stuff. You can fit over 700 copies of freaking War and Peace on the thing. Expandable memory would be nice just for the ability to fiddle with the contents without dealing with the Kindle itself, but the people for whom this is a huge deal are just looking for something to be indignant about.

But the one semi-legitimate complaint that you see commonly is the one about the file formats. I explained in my original Kindle review why most of these are stupid, but I’ll reiterate here.

The most common complaint is that the Kindle only supports a ‘proprietary’ file format. This is utterly ridiculous; it will read plain text files, which are about as unproprietary as you can get. People who complain that the Kindle ‘locks you in’ to any file format are either incredibly stupid, or they’re trying to spread FUD.

The Kindle will also read Mobipocket files; this is also a fairly open standard, if a bit harder to work with.

The Kindle will not read PDF files. This is kind of a shame, because everyone can create PDFs; but the complaints mainly center around some huge collection of pirated PDF e-books the complainer has, and that’s stupid. The pirate PDF editions you see are all formatted for either letter-size or A4-size paper, which just isn’t going to work on the Kindle; the screen isn’t large enough. And if you have the content in some other form — if it’s your own content, in other words — you can easily convert it to something that the Kindle will read — and the resulting file will be far smaller than a typical bloated PDF.

All of these complaints would go away if the Kindle could read HTML files (or, better yet, some kind of archive format where an HTML file and any supporting images were all stuffed into a zip file). All of the Kindle’s formatting capabilities would be accessible to anyone, using software they already have — and it would not require any actual capabilities that the Kindle doesn’t have already — just the ability for the web browser that’s already there to read HTML files from its internal storage rather than only from the network.

Mk. II Kindle Conclusions

The new Kindle is better than the old one in the most important way: the screen’s bit depth. The biggest improvement that’s still needed is entirely software: the thing needs better typefaces and the ability to read HTML files from its own storage.

A leading indicator, as the term is usually used, is a factoid that can be used to predict other factoids down the road. The business TV people love to talk about ‘durable goods orders’, for instance, even though I don’t think this is a very good indicator. People want fewer durable goods, and therefore you can infer from that that the people who work in the durable goods factories might soon be laid off, the suppliers of durable raw materials will soon see their orders shrink, etc., etc. The lack of orders leads the other things.

I don’t think that this makes much sense any more, largely because I think that the old models of the world economy are enough out of whack with current reality; I don’t think the models adequately consider the enormous increase in the velocity of information we’ve seen over the past ten or fifteen years. It used to be that by the time the durable goods factory (for instance) cut shifts, there was already a huge supply of durable goods in the pipeline, and the slowdown in demand had in fact been going on for quite some time. Today, a well managed company is going to make much quicker changes, so you’d expect to see more, smaller corrections rather than big, disruptive ones that come too late anyway.

And when the durable-goods factory does lay people off, they’re going to be able to find new jobs much more easily somewhere else thanks to the Internet. It used to be fairly complicated to get information about things outside about a 50-mile radius from where you were standing; now, it’s extremely simple. We understand this on some intellectual level, but our models of the world — both the informal ones we use to make sense of things, and the formal ones that the people on CNBC use to attempt to predict the future — have not fully integrated that information yet.

But this isn’t what I mean. I mean, I’m interested in what you might call micro leading indicators, the signs that indicate that a company or organization has just stopped giving a damn. These aren’t necessarily customer-service failures in any meaningful sense of the word, but things that suggest a plan or style of management that could lead to serious failure if the course doesn’t change. I saw a number of these yesterday at Costco.

The camera display there:

There are two problems here, really. The first, more subtle one, is that most of these cameras suck, and they’re almost all the same ones that have been sucking there for at least a couple months. Costco has never been a good place to buy cameras if you are particular about what you get, but it’s got worse in the past few months.

The more serious problem is that none of these cameras have any power. No batteries, no plugged-in power. The SLRs don’t even have lenses. This is a change from Costco’s previous approach. It means that Costco has spent the cost of one of each of these cameras for the purposes of a demo — but you can’t actually demo them because none of them do anything without power.

This thing is one of those dummy models. These are hateful to begin with, and something that you used to not see at Costco. I’m somewhat willing to forgive that because it’s a clock radio, and can’t be demo’d properly anyway without careful configuration, a radio signal, and an iPod or iPhone. This is a bit better than most dummy products, though, in that it will run through some kind of demo program if you hit the snooze bar. Unfortunately, it’s unplugged, and has been unplugged on this shelf for at least two weeks. You can see the power brick there next to it. The clock radio next to it is plugged in, but if you want to demo the one that works with the iPhone — sorry, Charlie.

They’ve gone to the trouble of plugging this TV in, but not of supplying it with a signal. This might not be such a terrible thing, since it’s a good price and since Sony hasn’t managed yet to destroy its reputation in the TV market as they have in so many others. This is a 1440×900 display, though, which is 16:10, not 16:9 as are most TVs these days. How well does this thing actually work? How does it deal with the scaling, and with the fact that TV signals are not quite the same shape as the screen? Impossible to say without buying it and taking it home.

Sylvania beat Sony again? Yes, in merchandising incompetence at Costco. This one isn’t even plugged in, and it certainly doesn’t come with a premium brand name. Sylvania started out over 100 years ago as a company that repaired burned-out light bulbs. They don’t actually manufacture consumer electronics; the company just licenses the name, in the case of TVs to Funai.

Making sure that all of this stuff is right — and particularly ensuring that the cameras have power — does take time and effort. Costco always has the manpower to have someone uselessly checking your card when you come in the door (that guy is really there to keep you from carrying stuff out the in door; you can’t buy anything without a membership card, so requiring you to have one to come in is pointless), and another to ‘make sure you have received all the items you paid for’ when you go out the door. Preventing shrinkage is very important to their business, and they staff accordingly. Catering to the customers is at least equally important, but since this is very hard to measure it is often one of the first things to slip.

If these leading indicators are accurate, I’d expect to soon see at Costco the following continuing indicators of the same trend:

• Safes on display locked so you can’t see the interior.
• Computers on display password-protected so they use electricity, but don’t allow you to do anything other than admire the case.
• More dummy electronics displays
• More slightly out-of-date products, or less desirable versions (e.g. only unpopular colors of iPods, maybe)
• Longer lines at the tills
• Eventually, sharply falling sales and membership renewals

The Virginia senate has effectively rejected two of our democratic governor’s pet causes: a doubling of the state cigarette tax, and a bill to ‘close the gun show loophole’, i.e. a law to require private individuals to conduct background checks on other private individuals to whom they sell guns (firearms dealers are required to conduct the checks no matter where they sell the guns).

In a Washington Post story about this headlined ‘Va. Still Holds Guns, Tobacco Dear’, a particularly obtuse state senator from suburban Fairfax County is quoted:

“Virginia still seems to be ruled by the gun lobby and the tobacco lobby,” said Sen. Janet Howell (D-Fairfax). “I think there are many members who are unwilling to oppose them for fear of retribution at the polls. And there also is a sort of traditional Virginia ‘past’ that is supportive of tobacco and guns.”

So: this is the fault of the gun and tobacco lobbies in Virginia; but members are unwilling to oppose those lobbies because, if they do, they expect retribution at the polls.

Or, in other words, many of the citizens of the Commonwealth of Virginia prefer that their state government not double the tax on cigarettes, and not make it harder to buy guns; if their representatives do not respect these wishes, the citizens are likely to vote them out. It seems to me that the vote, then, is representative democracy at work. Ms. Howell would apparently prefer a king.

That’s the headline on the Washington Post story: Local Police Want Right to Jam Wireless Signals.

There’s no good paragraph to excerpt in the story; the lead is about the Obamacade in the inaugural parade. A better lead would have been more clearly about the authority the feds apparently have to jam cell phone signals in certain situations.

Paragraph #2:

It is an increasingly common technology, with federal agencies expanding its use as state and local agencies are pushing for permission to do the same. Police and others say it could stop terrorists from coordinating during an attack, prevent suspects from erasing evidence on wireless devices, simplify arrests and keep inmates from using contraband phones.

I’m sure that this would be a useful item in the cop toolbox. But how many people think that the police wouldn’t abuse this power? The current cop mindset in the United States is one that leads to things like the Prince George’s County (MD) police storming the house of the mayor of a town in the county, killing his two dogs, and restraining him and his mother-in-law on the floor for hours because someone had anonymously sent a box of marijuana to his address — a box that was actually delivered by the police themselves?

The police wound up breaking down the door, shooting the dogs, etc. because, as they constantly claim, there might be a threat to their safety otherwise. ‘You never know what you’re going to find’ when serving a warrant, police spokesmen say again and again in justifying this kind of thing.

Maybe the reason they don’t know in advance what they’re going to find is that they’re not interested in finding out. From a recent Washington Post magazine story about the raid:

“The guy in there is crazy,” [Berwyn Heights, MD police officer] Johnson remembered a Prince George’s County officer telling him when he arrived. “He says he is the mayor of Berwyn Heights.”

“That is the mayor of Berwyn Heights,” Johnson replied.

Police work, ladies and gentlemen! You’d think, since the police spokesmen talk so much about the uncertainty and risk and so on that necessitates these home invasions, that they police would at least make the slightest attempt to figure out ahead of time what they’re going to be up against, no?

Apparently: no. The investigative work in this case, at least, didn’t even extend to having the slightest idea who their target was. Presumably they checked to see whether anyone living at the address had a criminal record, but I wouldn’t even be too sure of that.

There are 27 towns and municipalities in Prince George’s County, and the police in this case were so unfamiliar with the ‘community’ that they constantly talk about that they didn’t know that the guy was the mayor of the very municipality they were operating in.

Do you really think that a group of people so detail-oriented and respectful of civil authority would scruple to not just eventually jam all cell phone communications anywhere within, say, a 200-foot radius of every police car?

They wouldn’t do this at first, of course. But after they’d had jammers for a few years for use by SWAT teams and so on, they’d claim that the use of phones by people in cars during traffic stops represented some kind of threat, and they’d hook jammers up in all their cars. Cops would wear little portable jammers on their belts for those rare cases when they’re forced to venture more than 200 feet from their cars.

It’s worth noting that Brett Darrow’s 2007 recording of a St. George, MO police officer threatening to arrest him on phony charges was made by a camera that continuously uploaded its take. It’s a reasonable assumption that locally stored recordings of disputes between citizens and cops would generally be found to be corrupted, erased, or otherwise gone — just as the video cameras in police cars seem to malfunction at an astonishing rate, most often when there are suggestions that the recordings would tend to make the cops look bad.

In the District, corrections officials won permission from the FCC for a brief test of jamming technology at the D.C. jail last month, after citing the “alarming rate” of contraband phones being seized at prisons around the country.

“Cell phones are used by inmates to engage in highly pernicious behavior such as the intimidation of witnesses, coordination of escapes, and the conducting of criminal enterprises,” D.C. corrections chief Devon Brown wrote to the federal agency.

My main point isn’t about law-enforcement incompetence, but note that one of the needs they cite for this disruptive power is entirely the result of their own incompetence at keeping phones out of prisons. Either visitors are bringing them in, or their own guards are corrupt. Fixing either of those problems would be difficult, though: easier to just jam cell phones in the vicinity of the jail. That ordinary people would be inconvenienced by this isn’t important — they are mere citizens, after all.

The stated justification for all of this, of course, is movie-plot terrorism threats.

“When lives are at stake, law enforcement needs to find ways to disrupt cellphones and other communications in a pinpointed way against terrorists who are using them,” New York City Police Commissioner Raymond F. Kelly told a Senate panel Jan. 8. He also cited the Mumbai terrorist attacks, when hostage-takers used media spotters and satellite and mobile phones to help them outmaneuver police at hotels, train stations and other targets.

This, of course, ignores entirely the fact that much of the Mumbai carnage was the direct result of police incompetence. Witnesses to an early shooting at a railway station said that armed police there didn’t shoot at the terrorists.

It also ignores entirely the fact that if the police are known to jam cell phones, terrorists simply won’t depend on cell phones. They’ll use some other kind of radio that the police aren’t equipped to jam.

This would be an inconvenience for the terrorists — and I’m all for inconveniencing terrorists — but the police jamming would be an even greater ‘inconvenience’ for a citizen who might see something in the vicinity of an attack that might be useful to the police — but only if the citizen could call 911 on his cell phone.

This weekend, it was some kind of problem with my ‘credit balance’, even though I was trying to ‘purchase’ an app that was free. Now, it’s this:

Since ‘modifying’ an item takes, oh, the blink of an eye — or maybe a couple of seconds if you have to replicate databases — I cannot understand why a multi-billion dollar company is operating an online store — with custom client software, no less — in this way.

So, if you spend $3 million on a Super Bowl ad for beer, or corn chips, or candy bars, you advertise something that you sell every day — and something that you intend to go on selling every day until the end of time.

If you spend $3 million to advertise a particular car, you advertise something that you’ll only be selling for a year or so. But it’s a very expensive something as things go, and your ad also serves to strengthen your overall brand. Most Super Bowl ads are really about advertising the overall brand more than they are about selling a specific product.

So why do there seem to be as many ads for movies during the Super Bowl as for anything else? With the exception of certain ‘franchises’ of endless sequels — nearly all of which are incredibly bad and short-lived — there’s no ‘brand’ as such to establish in movies. Any particular movie is only going to be in the theaters for a few weeks; it’ll then be available as an expensive DVD for a few months; and then it’ll be in that giant bin at Wal-Mart, or on the Amazon four-for-three deal.

If it turns out to be a fantastic movie, of course, it’ll be in the theaters and on the regular-price DVD rack longer; but then this would be the case even without the Super Bowl ad. But these movies aren’t going to be fantastic; one of them is another The Fast and The Furious sequel, for Christ’s sake. And all of these movies being advertised seem to be scheduled for release in about five months, by which time most people will have forgotten about the ads anyway.

How on Earth does this pay off? Must be something about Hollywood’s Chinese accounting.

In the Washington Post this morning, we learn that no less august a personage as Joe Biden is going to head up a new Task Force.

Biden made his first prominent White House appearance Friday with the launch of the Middle Class Working Families Task Force, billed as a “major initiative targeted at raising the living standards of middle-class, working families in America.”

You know, it seems that very recently, these same people were telling us that the living standards of middle-class Americans were Destroying The Earth, and that we all had to Cut Back or the Seas Would Rise etc., etc.

There are undoubtedly some people in the United States who could use a boost in their standard of living. And there are some others who really couldn’t have a boost in their standard of living, barring the discovery of hitherto unknown loopholes in the laws of physics or the invention of some new and wonderful cocktail; these people have a standard of living that is already as high as is possible given the current state of technology.

In the middle are the middle-class, the people for whom an improvement in their standard of living is theoretically possible, but who are not going hungry, or cold, or without a car or TV or boat (if they want a boat) or, really, pretty much anything else this side of stained-glass bathrobes and world’s fattest racehorses. The middle-class American lifestyle is the envy of pretty much everyone in the world — excepting only the contemporary super-rich, and even they might see some things to be admired in it.

But it’s not enough for Joe Biden.

Putting aside political cynicism, the only conclusion I can arrive at is that they are talking about something else. ‘Middle class’ as used by Biden et al. does not mean what it means when used by humans.

I think that by middle class they mean what used to be called the working poor. (The working poor used to be called just the poor until people started to notice that, in a hell of a lot of cases, the poor were poor because they didn’t do any work. The majority quite rightly doesn’t feel a lot of sympathy for such people, so people who worked but didn’t make much money started being called the working poor.)

The chief clue is in the name of the Task Force: the Middle[-]Class Working Families Task Force. Barring people who are temporarily unemployed, are there any middle-class people in the United States who don’t work? Presumably, if you don’t work but are able to live off your investment income, you are rich (or at least retired); if you don’t work but are not able to live off your investment income, you are poor. If you do work but are not able to live off that income, you are the working poor.

It used to be that everyone else was middle-class. Now we have a new thing, this ‘working’ middle class. Presumably as the Task Force spins up, the meaning of this phrase will become a little less opaque.

Sea birds tend to congregate in this too-big parking lot in Warrenton, VA — at least 70 miles from the nearest salt water. There are about six of them on top of each of the light poles.