Category Technology

iPad Bag Recommendation

So everyone seems to be having trouble finding an appropriate bag for their iPad. I certainly haven’t yet seen anything that looks good. Bags meant for laptops are too big (and my experience is that many of them actually depend on the presence of a laptop of a certain size to hold their shape), and everything else is essentially a purse.

I imagine that this is a problem even for the ladies, because the demographic that will tend to own an iPad right now is the demographic that’s likely to be carrying a purse that’s far too small for the thing.

My solution, for now at least, is the Healthy Back Bag. Mine is actually from LL Bean, but they don’t seem to sell this particular model any more. I have the microfiber bag in black, size Small. It does not look like the iPad should fit in there, but it does, and without squeezing.

There are another four pockets outside and seven pockets inside that are good for headphones, cameras, my little folding keyboard, pens, notecards, etc., etc., etc.

Click on these pictures to see bigger ones.

bag_inside_1.jpg bag_inside_2.jpg bag_inside_3.jpg
bag_with_ipad.jpg bag_with_ipad_2.jpg bag_ipad_inside.jpg

The iPad itself first goes into a homemade leather sleeve that Nicole made me. This provides some minimal cushioning from some thin foam on the flat parts, and from the seams on the edges.

The Healthy Back Bag isn’t what you’d call handsome, but it’s extremely light weight (when you’re going to be carrying something that’s itself fairly light, it pays to not waste most of that savings on the bag itself). Because it’s not a purpose-built computer bag, it also has the advantage of not shouting EXPENSIVE GIZMO to everyone who sees it. This may not actually be an advantage for you, depending on your outlook.

My original plan was just to use this until some decent Official Solution came on the market, but the improvisation is working so well that any O.S. will have to be very good indeed to pry me away from this.

Two Confusing Things About Publishing

From a recent, much-linked New Yorker article on the e-book business:

Good publishers find and cultivate writers, some of whom do not initially have much commercial promise. They also give advances on royalties, without which most writers of nonfiction could not afford to research new books. The industry produces more than a hundred thousand books a year, seventy per cent of which will not earn back the money that their authors have been advanced; aside from returns, royalty advances are by far publishers’ biggest expense. Although critics argue that traditional book publishing takes too much money from authors, in reality the profits earned by the relatively small percentage of authors whose books make money essentially go to subsidizing less commercially successful writers. The system is inefficient, but it supports a class of professional writers, which might not otherwise exist.
  1. There are not many businesses where you can be wrong 70% of the time and stay in business.

  2. Returns — the publishers’ biggest expense are entirely eliminated by e-books. That must be why they insist that e-books should cost as much as, or more than, hardcover books.

Review Of An iPhone-Killer

I’ve complained in the past about tech reviews; that it’s basically impossible, no matter how awesome a reviewer you are, to write an accurate review of a complicated product without using it — really using it with (in the case of a computer etc.) your real data — for an extended period.

Car magazines understand this, which is why they have their long-term test cars. I gather that if you work for a car magazine, you are in the strange position of being fantastically knowledgeable about the what’s available in the car market, what’s good and what’s not, etc., etc. — but that you don’t actually ever have to purchase a car of your own. You’re always either driving a car for short-term review, or you’re driving one of the magazine’s long-term cars.

The long-term reviews shed proper light on fiddly things like the trunk sill being too high, or the air conditioning sapping too much power when driving uphill at altitude in the summer, or that the cup holders are crap. These are things you wouldn’t notice at all in a weekend, but that loom up to become the defining characteristics of some cars after long enough.

When it comes to computers or any sufficiently complicated phone, or camera, or anything like that, it’s all fiddly things. You can’t really judge anything except in a long-term review.

The Apple retail stores are better than any other computer store I’ve ever seen; the computers all work, they don’t have screen-saver passwords that keep you from being able to use them, and they’re loaded with actual software that should, in theory, allow you to evaluate at least how zippy the computers are. But you can’t. Or I can’t, anyway. Without my data, and my actual work, I can’t tell you a damned thing about the performance.

This is particularly a problem with smartphone reviews. Nobody reviewing a smartphone uses the device under review as their only phone; my guess is that the only calls anyone actually receives on review phones are those that are placed by their secretaries or friends specifically as test calls. You’re not generally going to keep your real calendar on the thing, because you already have a phone for that. You’ll probably copy your address book onto the review phone, but since you’re placing most of your regular phone calls on your regular phone, you might not notice a lot of little problems on the review phone.

There’s nothing really wrong with this; it’s easier to use a review car for a week than it is to use a review phone, after all. And the better reviewers have a lot going on; it wouldn’t be reasonable to expect them to throw their whole portable-communications setup into turmoil for every successive phone they review.

So most tech reviews, even those written by people who are genuinely making their very best effort to do right by their readers, suck. Such is the nature of the product.

Today, Jamie Zawinski offers his opinion of the Palm Pre after using it for a while as his actual everyday phone.

In part, he says:

Peformance is a joke.

Seriously, it’s comically bad. The speed of this phone is truly pathological. It’s horrible across the board, but some of the most egregious examples:

  • If the Calendar app is not running, it takes 10-15 seconds to get from “I clicked on the Calendar icon” to “I can see today’s events”. And then, switching from the display of one day to the next takes 2+ seconds (and it doesn’t buffer swipes, so you have to keep trying). It’s embarassing when I’m talking to someone and they ask me about availability and I have to say, “I’ll tell you in a little while, once my phone wakes up.”

  • If a call comes in, the phone starts ringing, and I can answer and talk to the caller, but most of the time it takes another 10 seconds before the Phone application’s UI comes up! So if it’s from the front door and I have to press a button to buzz someone in, I have to either hope the app starts responding before the caller hangs up; or I have to slide out the physical keypad and pray that it buffers the keystroke. Trying to answer the door feels like a game of whack-a-mole.

  • If I want to take a photo (for example, of the license plate of a hit-and-run) getting from “I clicked on the Camera button” to “I have taken a photo” takes almost 20 seconds. If I want to get all the way to “I have reviewed the photo, and can tell that it came out ok”, that takes more like 40 seconds.

The Pre is the most credible ‘iPhone killer’ currently on the market.

Yet Another iPhone Killer Identified

Michael Arrington, of all people, writes in The Washington Post — or maybe just the Post‘s website, since it’s impossible to tell whether something was actually printed in the paper or not:

Verizon and Motorola finally lifted the curtain on their new Droid Android phone yesterday. Make no mistake, this is Android’s flagship product, and the first phone that will pose a significant threat to Apple’s iPhone.

I’m sure the Droid will be an excellent device. I even suspect that Arrington may be correct in that the Droid will finally provide at least some competition for the iPhone.

His statement would have more credibility, though, if clueless reviewers hadn’t dubbed damned near every single awful smartphone released since June 2007 an ‘iPhone killer’.

Crunchgear:
200910180844

ZDNet:
200910180844-1

PMP Today:
200910180845

Telegraph:
200910180845-1

Gizmodo:
200910180845-2

Businessweek:
200910180846

ZDNet again:
200910180846-1

Unwired View:
200910180847

Today Show via MSNBC:
200910180847-1

CBS Evening News:
200910180848

IntoMobile (note that this was written before the original iPhone was released):
200910180849

Betanews:
200910180849-1

SearchMobileComputing:
200910180850

I do have to wonder whether the recent FTC web-endorsement-payola B.S. has some of its roots in this kind of thing. Since before the original iPhone was released in June 2007, it seems like every new smartphone — even ones that turned out to be colossal stinking failures, like the Blackberry Storm — has been proclaimed ‘the iPhone killer’ by someone.

In many cases, the manufacturers and carriers have even been claiming that their device is an ‘iPhone killer’. Note that Verizon did this two years ago with the LG VX10000 Voyager. Now, I understand that Verizon is in the business of selling these things and thus needs to present them in their best light and to boast about them as much as possible. But anyone who would seriously consider that thing a rival to the iPhone should have his head examined. The hardware was neat, and the Verizon network was and is vastly superior to the AT&T network that iPhones are locked to in the United States. But the software sucked, just as it still sucks on just about all phones. Nobody seems to take notice of this.

There’s definitely a lot of room for improvement in the iPhone, and I’m certain that something will eventually come along that’s better — though to be honest the next better-than-the-iPhone device is likely to be the iPhone Mark IV. These reviews by iPhone-hatas (I assume they’re iPhone-hatas because they seem to see everything through the lens of iPhone-killing) don’t help matters, though: if it hasn’t happened already, the public eventually will learn that ‘iPhone killer’ nearly always means phone with neat industrial design and terrible software and UI.

The Kindle After Almost Two Years

I’ve written about the Amazon Kindle before here, on a number of occasions. I think, but cannot be arsed to check, that I have pointed out that the Kindle is a gadget designed for long continuous use, and that a lot of its reviews — which involve a reviewer who might not necessarily be a reader poking at the thing for a few hours — reflect this.

Nicholson Baker’s recent long-form Kindle review in the New Yorker illustrates both approaches in the same article.

He starts out with nitpicking:

The problem was not that the screen was in black-and-white; if it had really been black-and-white, that would have been fine. The problem was that the screen was gray. And it wasn’t just gray; it was a greenish, sickly gray. A postmortem gray. The resizable typeface, Monotype Caecilia, appeared as a darker gray. Dark gray on paler greenish gray was the palette of the Amazon Kindle.

This is a legitimate criticism of the Kindle; the screen isn’t white and the type isn’t black. Part of this is undoubtedly due to the fact that the device itself is white, and that the screen’s ‘white’ feels darker than it really is by contrast.

But in another way it’s not a legitimate criticism, because the Kindle’s purpose is not perfect color fidelity. The screen is perfectly readable, which is the only point. This tendency toward a Failure To See The Point is the main reason I titled my original Kindle post ‘The iPhone and the Kindle’. The iPhone, which was released a few months before the Kindle, was (and in some cases to this day is) conspicuously missing a few features that are almost universal in cheaper phones.

For example, in the US, you still can’t send or receive MMS messages with an iPhone. You can’t transfer contact information (or, really, anything else) using Bluetooth. You can’t transfer, say, contact information from one iPhone to another — which doesn’t sound like that big a deal until you need to do it. You can’t add storage to the phone by sticking in a cheap memory card. When the iPhone was first released, you were limited to only the pre-loaded Apple applications, and Apple’s strong implication at that time was that this was all that would be allowed forever. The camera sucked. Many of the initial reviews focused on these things to the exclusion of the fact that the iPhone was, at the things it did do, at least a couple orders of magnitude better than anything else.

And so it is with the Kindle. I’ve made this point before, but I’ll make it again because it’s so important. The whole point of the Kindle is the electrophoretic screen. That’s it. You can read electronic books on your computer screen (or iPhone screen) ‘for free’, but that isn’t as good as reading them on the Kindle’s screen, and with the Kindle’s impressive battery life. This fact continues to be lost on most of the reviewers, because they dink around with the thing for a little while and conclude (correctly) that the navigation controls are bad.

This is only a tiny bit less irrelevant than is pointing out that the iPhone is at best a poor hammer. The Kindle has only one important navigation control: the ‘Next Page’ button. The Kindle has only one important mode of use: reading page after page of text. Everything else it does is hung on the side, and the device isn’t optimized for it.

By the end of his New Yorker article, Baker has seen the light, at least a bit:

Then, out of a sense of duty, I forced myself to read the book on the physical Kindle 2. It was like going from a Mini Cooper to a white 1982 Impala with blown shocks [compared to reading the same book on the iPhone]. But never mind: at that point, I was locked into the plot and it didn’t matter. Poof, the Kindle disappeared, just as Jeff Bezos had promised it would. I began walking up and down the driveway, reading in the sun. Three distant lawnmowers were going. Someone wearing a salmon-colored shirt was spraying a hose across the street. But I was in the courtroom, listening to the murderer testify. I felt the primitive clawing pressure of wanting to know how things turned out.

The Kindle disappears. There you are.

However, after almost two years of using the Kindle, I have come across a few things that I believe are genuine flaws.

1. The Kindle is too modal. The Kindle has two gross states: awake, and asleep. When the Kindle goes to sleep, the text on the screen is replaced (at a small cost in battery power) with a ‘screensaver’ image (though the screen doesn’t need to be ‘saved’ in the way old computer CRTs did) featuring, usually, some famous author. You can put the Kindle to sleep manually by frobbing the switch, or if you leave the thing alone it’ll go to sleep on its own after ten minutes. This wait-before-sleep time is not configurable by any normal means.

After long enough, this gets annoying, because it takes a couple seconds for the Kindle to wake up and put text back on the screen. Since the Kindle doesn’t consume any appreciable power while it’s just sitting there idle and waiting for you to turn the page, the whole purpose of the screen saver is to show you that the device has gone to sleep and locked the buttons. It would be far, far better to have a physical switch that locks the buttons while leaving the screen alone. The user could check the current mode by looking at the physical switch; power wouldn’t be wasted putting up a bad picture of John Steinbeck; and I wouldn’t have to wait for the thing to re-draw the screen before starting reading again.

2. The Kindle’s wireless access is too modal. You can turn the wireless service on, or you can turn it off. Leaving it on results in the batter being drained much more quickly — it dies in something like 2 days vs. a week with the wireless service off. There’s no real advantage to having it turned on while you’re just reading a book.

But if you read books on both the Kindle and an iPhone — as I have recently started to do — it is vitally important that the Kindle have the opportunity to upload your current position in the book when you put it down. When you fire up the iPhone Kindle application, it syncs your position and you can pick up where you left off. This is brilliant, and it is the entire reason for a number of Kindle book purchases I’ve made recently.

For most of my Kindle-using life, I’ve been loading it with text I have obtained elsewhere; this is easy to do, and it’s indistinguishable, for the most part, from official DRM’d Amazon Kindle Editions.

But only the official DRM’d Amazon Kindle Editions will sync your position between multiple devices. This is a hell of a feature, a real added value that helps justify the bafflingly high prices the publishers want for some of these e-books.

The problem, though, is that this adds another step that gets in my way. I keep the wireless service turned off most of the time, but I need to turn it on for a little while when I pick up the Kindle (to pick up updated positions from iPhone reading), and when I put the Kindle down (to send updated positions for the benefit of the iPhone). That’s nuts. The Kindle’s wireless service should have three states: ‘On’, ‘Off’, and ‘Off, but connect after I’ve read something and send the position, and connect when I pick up the device to read again to check for updated positions, and connect once a day or so to check for new content’. ‘On’ already approximates this — the radio isn’t turned on at full blast the whole time — but the battery life suggests that it spends a lot more time powered up than it really needs to for my purposes.

3. The Kindle needs to look better. I was a defender of the appearance of the Kindle early on, but the plasticky nature of the thing (I currently use a Kindle 2) is starting to wear on me. I still believe that the extremely mundane look of the Kindle was and is deliberate; the idea is that it shouldn’t look so gadgety. I think that my change of opinion is related to my use of the Kindle and the iPhone Kindle app together somehow; the Kindle has moved from being just a display device (where the plastic business didn’t bother me) to being a futuristic reading device out of a science-fiction novel. The problem is that the ‘white and plasticky’ part gets in the way of the ‘science fiction’ part. It’s time for the Kindle to start looking at least a bit more like the sci-fi marvel that it is.

I fully expect the upcoming announcement or release of Apple’s now rumored-to-the-point-of-near-certainty ‘tablet’ device — which will, of course, look like something from the future — to produce a lot of ‘Kindle Killer’ type stories, which will all be written by non-readers and predicated on a misunderstanding of the screen technologies.

Slingplayer WTF

So I launch SlingPlayer, and it tells me that there’s an update available.

Picture 1

That is, the OS X version of SlingPlayer tells me that there’s an update available. I click on ‘yes’, I would ‘like to start download now’.

But the download doesn’t start. Instead, I get a web page that asks me to select what I want to download. Shouldn’t SlingPlayer know this, and pass the information along?

Picture 2

Next, it doesn’t know what OS I want this for, either:

Picture 3

Clicking on the ‘SlingPlayer for Mac’ button actually starts the download.

Why? Why on Earth? The effect of pushing the ‘Yes’ button in the application is just to open a certain URL in the default web browser. Why is that URL not the one that’ll start the download? Or at least a page that presents a single button that allows me to download SlingPlayer for OS X, which is the only thing I can possibly download in order to upgrade the current installation. The application doesn’t ask whether I want to download a different SlingPlayer; it doesn’t give me any information about whether my SlingPlayer for Windows or SlingPlayer Mobile (which is itself actually several different applications depending on your device) is out of date; it asks whether I want to upgrade the software I am using now, and then doesn’t bother to send me directly to the appropriate place for that.

This has been this way for several versions, too. Part of the problem is undoubtedly that most people within the Sling Media operation — and certainly the people who actually write and test the software — never upgrade via this path; they’re using beta versions or pre-release golden masters or something, and are not likely to ever be unaware that a new version is available.

It’s a very small thing — two extra clicks — but because fixing it would be a matter of spending ten seconds, once, to change the URL stored in the application, it suggests an inattention to detail that’s apparent elsewhere in the application, too. Like the volume control, for instance:

With the system volume turned up to about 75% and the speakers set to half way — the way I listen to pretty much everything else — this is a comfortable volume in my office:

200907160756

See that? Notice that none of the volume bar is visible to the left of the slider The application says that this is 6%.

This (30%) is ridiculously loud:

200907160757

The result is that if you want the TV murmuring in the background, you have to do it by turning down the system volume, which isn’t good if you want other sounds to be audible.

Even Dumber Than Usual

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.

Kindle Typographical Addendum

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:

Kindle Decent Typeface1 Kindle Decent Typeface2

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.

New Kindle

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:

2Bit

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:

4-Bit

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:

4-Bit-Dithered-1

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:

Dither-Closeup

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:

T-No Anti Aliasing

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:

T-Anti Aliased

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

T-Small Anti-Aliased

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:

Phrase Antialias

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.

Original Kindle Text New Kindle Text

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:

Caecilia

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.

iPhone Location Error

A lot has been said about the iPhone’s ability to locate the user; the new 3G iPhone actually uses GPS to do this, while the original iPhone just uses triangulation from cell-phone towers to figure out where you are.

The triangulation method should work reasonably well for most purposes (i.e. ‘what town am I in?’ at the very least), but my experience lately has been that this isn’t the case.

The other day, I hit the Locate button and got this result. Click on the picture for a bigger version:

Iphone Location Error

Where I was standing, the terrain is such that there’s no signal at all from the south and west. Where I was standing, the elevation is 700 feet, and the mountains are about 2300 feet. You can see the problem (if you squint) in this view looking roughly south. Click for a bigger version:

Terrain View

There’s also a low ridge between the airport (where I was standing) and the town of Front Royal itself. It’s not enough to really notice, but it’s enough to block cell phone signals. My guess is that the phone could get only one signal, and that this was from a cell site on or near Mouth Weather, which is just NE of the identified (wrong) location. All the system could tell was that I was somewhere on a line roughly 210 degrees from that tower.

The lesson: cell-phone triangulation location information can be ridiculously inaccurate when in areas with interesting terrain and poor coverage. More importantly and less obviously, the locator system appears to be unable to reliably determine when this is the case.