I’m : a programmer, writer, podcaster, geek, and coffee enthusiast.

Clean installs on Macs

Garrett Murray is doing a clean install for the first time in a while:

I’ve been using the same core install of OS X since April 30, 2005, the day I received my copy of 10.4 Tiger and did a fresh install on my Power Mac G5. Over the past six years I’ve installed the two subsequent releases of OS X as upgrades (Leopard and Snow Leopard), and planned to do the same with 10.7 Lion when it arrived this month.

I’ve been meaning to write a similar post because I’ve followed almost the same path. I’m still using the same OS X installation, migrated between Macs and upgraded to new OS versions, since from my first Mac: a PowerBook G4 running 10.3 in 2004.

The title bar still says this when I open a new Terminal window. I left Vivisimo in 2006.

I have applications that haven’t been updated since 2006. I probably haven’t launched them since 2007. I have 60 hidden folders (whose names begin with dots) in my home folder from various apps storing preferences there. The Unix subsystem has been horribly cluttered by various package managers and custom installations of PHP and MySQL over the years. Application Support has 116 folders, 58 of which haven’t been touched since 2009 or earlier.

And I just cleaned out a lot of old applications last month, apparently poorly.

Such a long-running installation is uncharacteristic of my pre-2004, PC-using self: I was one of those guys who reinstalled Windows every 6 months just to screw around, set things up differently, or get a perceived speedup that may not have actually been there.

But then I (barely) graduated from college, got a full-time job, and lost the drive to keep screwing around with my OS. I just wanted to get things done in my precious few free hours, and fiddling with my setup began to feel like wasted time. This shift in my priorities is what has kept me so happily addicted to Apple products.

Generally, Mac OS doesn’t need clean installs. Disk cloning, Migration Assistant, and each new OS’ setup procedure all work very well to upgrade your system in place. This is implausible to former Windows users, but trust me, it really does work. That’s how I’ve carried the same installation across five Macs, three major OS releases, and two processor architectures.

I could keep it going to Lion, my installation’s fourth major OS release. But, like Garrett, I’m going to start fresh this time, for a number of reasons:

I just need to find a good time, and enough motivation, to spend a few hours doing all of this some evening instead of anything else that I’d rather be doing.

Review: Puerto Rico HD for iPad

I’m a huge fan of the Puerto Rico board game: it’s one of the best-known (and one of the best) European-style board games, a category finally becoming popular in the United States with the success of Settlers of Catan.

Unlike most games popular in the U.S., this genre is known for relying much more on strategy and less on luck. Settlers of Catan is actually more of a hybrid in this way: I’d say it relies on luck about as much as Monopoly does, in that even a great strategy can be sunk from a streak of bad luck with the dice.

Puerto Rico is almost pure strategy, with only two relatively inconsequential random elements — the plantation tiles available in each Settler phase, and the initial play order (who goes first) — and the game effectively minimizes and balances the impact of these. It’s no surprise, then, that it dominated BoardGameGeek’s top-ranked list for most of the last decade.

It requires at least three people to play, so a geeky board-game couple like us can only play it as often as we can convince someone to come over and learn it. Everyone we’ve taught to play1 loves the game, including non-geek friends, parents, and siblings, but there are a lot of pieces and a lot of rules. It’s of similar complexity as Settlers of Catan and is therefore best taught virally: you play with someone who knows it and explains it as you go. That’s how I learned (thanks, Andrew!), and I can’t imagine three people sitting down and trying to figure it out from scratch.

As soon as I learned that there was an upcoming Puerto Rico iPad game, I was thrilled: I could finally play this awesome game more often than the twice a year or so that I could convince enough people to play. Plus, the iPad is an excellent board-game platform, so why not port an excellent board game?

I’ve been testing an advance copy of Puerto Rico HD this week, and it’s now available in the App Store. In short: Fans of the board game are going to love this.

Click for big. Or see later in the game.

The board game’s assortment of cards and tiny pieces makes for a huge interface challenge. While there are some small things I would have done differently, Codito Development did a pretty good job with this. It’s clear that they really care about doing the game justice: it’s nothing like the half-assed licensed ports that we usually see from well-known board-game brands.

I was worried that such a complex game wouldn’t have a very good AI, but this AI is excellent. I win regularly against the Medium opponents, but the Hard AI players beat me often. It’s the perfect balance of AI skill without making the game too easy or frustratingly difficult. Single-player is therefore very satisfying, and I can play a speedy game in about 15 spare minutes.

I didn’t have a chance to try the online multiplayer via Game Center, but while it doesn’t support asynchronous turn-based play with multiple simultaneous running games (like Words With Friends), the developer has hinted that it might get this feature in the future.

Tiff and I did play some hot-seat games (both of us taking turns in front of one iPad), and it works extremely well, with the ability to mix any number of human and AI players.

A nice little trick makes hot-seat play exceptional: in the player setup screen, you can tell it which side of the iPad each player is sitting on, and it automatically rotates the interface to face each player on their turn. (Only the two landscape orientations are supported.) That way, if you’re sitting at a table, you can lay the iPad flat in the middle and play comfortably without rotating it constantly. And you can comfortably watch your wife decimate you by buying a wharf and a harbor and shipping out incredible quantities of corn.

Playing the physical board game is an hour of tangible fun with friends or family; playing the iPad game is 15 minutes of amusement with any number of friends (including zero), no setup, and no cleanup. Neither replaces the other. One of the best uses for the iPad game is improving your skills and discovering new strategies to make your real-world gameplay more interesting.

I love that this game is finally available for iPad and we now have that choice.

My only hesitation is that I don’t think I could recommend the iPad edition to people unfamiliar with the board game. Maybe it’s just me2, but I don’t think this would be an easy game to figure out on your own in an app, through no fault of Codito or Ravensburger. The game does have a built-in tutorial, strategy suggestions, and complete instructions, though, so I’m curious: if you’re able to figure out how to play from the app with no previous experience with the board game, please let me know.

If you’re familiar with the board game, or you’re confident that you could learn it from the app, get it from the App Store now. It’s a great, solid iPad adaptation of the best board game I’ve ever played.

  1. It’s always a bit uncomfortable teaching new players about taking these brown “colonists” off the boat and putting them to work on our tobacco plantations without paying them so we can ship crops to the mainland.

    (Settlers of Catan isn’t much better. What’s the only black piece in the game?) ↩︎

  2. I still can’t figure out Carcassonne from its widely-praised iOS app alone. I just don’t have the patience to learn a complex board game from its computerized version. ↩︎

Review: Coffee Joulies

My Coffee Joulies just arrived from the first batch of shipments to its Kickstarter backers. The little card that comes with them gives a good description of what they’re supposed to do:

Place Joulies in your cup before pouring your coffee and it will be ready to drink three times sooner and will remain hot twice as long.

See their original Kickstarter page for information as well.

How hard could it be to photograph a bunch of black, white, and highly reflective objects on a reflective black surface with a 100mm lens indoors?

They look great.1 They’re also much larger than I expected, although that’s not a bad thing.

These feel like extremely high-quality items, and they also don’t impart any metallic taste that I can detect. They seem very well-made.

Their site details their intended thermal behavior:

  • First, they absorb energy from your coffee when it is too hot, cooling it down to a drinkable temperature three times faster
  • Next, they release that stored energy back into your coffee keeping it in the right temperature range twice as long

As we discussed on my podcast when they were announced, they contain a phase-change material (PCM) with a melting temperature of around 140 degrees Fahrenheit, supposedly the ideal coffee-drinking temperature. Wikipedia explains how it works:

Initially, the solid-liquid PCMs behave like sensible heat storage (SHS) materials; their temperature rises as they absorb heat. Unlike conventional SHS, however, when PCMs reach the temperature at which they change phase (their melting temperature) they absorb large amounts of heat at an almost constant temperature. The PCM continues to absorb heat without a significant rise in temperature until all the material is transformed to the liquid phase. When the ambient temperature around a liquid material falls, the PCM solidifies, releasing its stored latent heat. … They store 5 to 14 times more heat per unit volume than conventional storage materials such as water, masonry or rock.

Sounds good. Let’s see how well it works.

I first tested them in a common scenario: putting room-temperature Joulies in room-temperature mugs, pouring matching amounts of boiling water into a control mug and the Joulies-equipped mug of each type, and measuring the water temperatures every 2 minutes until both mugs in each pair fell below 110ºF (which I consider too tepid to comfortably drink).

This is not lab-precision equipment, and these were not lab conditions. I couldn’t, for instance, measure both cups’ temperatures exactly simultaneously, so I alternated the thermometer between mugs until its temperature stabilized in each (about 10 seconds). I poured the water in the same sequence with the same delay between the two mugs, so any skew from this should be minimal. Regardless, the measurements I’ve taken should be considered approximations.

First, the results from my favorite insulated mug for road trips, the Contigo 16-ounce travel mug:

Keep in mind that the lids were off for the constant thermal probing, so they lost heat to the air much faster than they would if they were sealed.

You can see the effect of the Joulies: the temperature in their mug dropped faster than the control mug, then slowed its drop noticeably at around 140ºF, exactly as designed. The temperatures soon started dropping at about the same rate, but the mug with the Joulies was slightly warmer.

The control mug reached 140º at 43 minutes, and the mug with Joulies reached it at 37 minutes.

The control mug fell below 110º at 111 minutes, and the mug with Joulies fell below at 135 minutes.

The effect is present, but weak. And it’s worse in ceramic mugs, where much more heat is lost to the mug’s walls, the surface it’s sitting on, and the surrounding air:

The control mug reached 140º at 13 minutes, and the mug with Joulies reached it at about 12 minutes. Both mugs fell below 110º at 37 minutes.

This wasn’t looking good for the Joulies, but to give them another shot, I created a favorable scenario: after everything reached room temperature again, I ran another test using all five Joulies in a ceramic mug, and warmed the Julies in some hot tap water first.

The results were more amplified, but ultimately didn’t help the Joulies much:

The control mug reached 140º at 11 minutes, and the mug with Joulies reached it at 7 minutes. The large gap at the start is most likely due to the huge mass of Joulies absorbing so much of the initial heat.

The control mug fell below 110º at 31 minutes. The mug with Joulies fell below at 33 minutes. (I expect these were worse than the previous ceramic test because of the smaller water volume that would fit in the mug with all of those Joulies.)

I could do more tests with different conditions,2 but I honestly don’t want to spend another four hours to reinforce what seems pretty clear already: Coffee Joulies do work, but their effect isn’t very strong, and it’s nowhere near their claims that the drink “will be ready to drink three times sooner and will remain hot twice as long.” In fact, the effect is barely noticeable.

I imagine they’d fare well in an all-day test in a vacuum mug or bottle, but the effect is going to be similar to my insulated-mug test: an improvement, but not by much.

Their website is already playing down their benefits with regular mugs with statements like this:


But that statement is just as true without Coffee Joulies: good vacuum mugs like my trusty Contigo already keep liquids at hot (or cold) temperatures for far longer than most people expect.

While Joulies may help a bit, they’re probably not worth their cost to do so. (I don’t know what their final retail cost will be, but their cost on Kickstarter was $40.) And in addition to the cost, there are other downsides to using them:

I wish Joulies had more of an effect, especially in the common mug-on-a-desk scenario. I was excited for this project since I saw it on Kickstarter, and the creators seem like nice, hopeful guys. I honestly feel bad for them — I really wanted their product to be great, but I can’t recommend it.

If anything, this should convince you of the value of a good travel mug.

  1. Yes, their instruction card is printed entirely in Comic Sans. But I won’t hold that against them because the Joulies look so good. ↩︎

  2. Some ideas for more tests: using Joulies that had sat in boiling water for a while to be as hot as possible; testing with two good vacuum bottles over 8 hours; comparing the performance of Joulies to equal masses of rocks or metal. ↩︎

An iPad success story

In a recent visit with my mother, she showed Tiff and me the plans for a home renovation and was trying to figure out how to lay out the kitchen. We had some ideas, but it was hard to communicate them properly with hand-waving and our terrible drawing skills. But we ran out of time that morning and had to take a day trip an hour away.

Tiff was driving, so I brought my 3G-equipped iPad, searched for home-design apps, found a few decent-looking ones, bought Home 3D, figured out how to use it, and started recreating my mother’s new floor plan. All from the car, in the middle of nowhere in upstate New York.

When we got back to her house, I showed her this:

The iPad is only for consumption.

Then we 3D-“walked” through it. And it was awesome. She showed it to the contractor, and it has become the basis for the new design of her home.

The app isn’t amazingly great. Many parts of the interface are unintuitive, sloppy, or ugly. In addition to its $8.99 price, nearly all of the good furniture (including some essentials, like cabinets) is only available via In-App Purchase for a few more dollars, so it’s hard not to feel nickel-and-dimed. There are very few options for each item — you only get one or two choices for most objects. And all of the data is stuck in this silo, with only the ability to export images.

But this effort was still a huge success, because even with all of those limitations, it did everything we needed it to do.

I’ve had laptops and cellular internet connectivity for 7 years, but I never would have done something like this before. Why?

The computing revolution brought on by iOS, the hardware, and the App Store ecosystem is a bigger deal than we realize.

Why software patents are not fixable

I’m not a patent lawyer. I’m not even a lawyer. I’m just a software developer, and like every software developer, I’ve probably unknowingly infringed upon hundreds of patents while routinely doing my job.

In my efforts to educate myself on the patent system, I’ve learned about the requirements for getting a patent, and what sorts of ideas are patentable. One of these requirements is novelty:

An invention is not patentable if the claimed subject matter was disclosed before the date of filing, or before the date of priority if a priority is claimed, of the patent application.

A patent must also not be obvious to a person of ordinary skill in the field that the patent pertains:

A patent may not be obtained though the invention … if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.

But as we’ve seen, time and time again, the U.S. Patent and Trademark Office enforces these (and other) rules extremely poorly, resulting in thousands of patents being issued that most people knowledgeable in their field would immediately recognize as invalid.

And that’s not new. Bogus patents have been issued for decades.

The USPTO has repeatedly shown that they cannot and will not do their job to prevent most invalid patents from being granted. In the field of software, their negligence seems especially egregious.

With the rise of patent litigation due in no small part to the “we don’t sue people” Intellectual Ventures and their shell companies (that only sue people), the number of patent applications is likely to increase dramatically over time. If the USPTO is granting bogus patents because they’re just slipping past overworked patent examiners, it’s only going to get worse.

Invalid patents aren’t just funny government slip-ups. (“Oh look, someone patented toast and linked lists! Stupid patent office!”)

Since the economics of civil lawsuits, especially patent lawsuits, prevent most cases against small defendants from ever getting near a court, the potential cost to society of issuing an invalid patent is massive.

If someone threatens your small business with a patent lawsuit, it doesn’t matter whether the patent is valid. Because for you to prove that it’s invalid would take far more time and money than you probably have. The only sensible course of action, the path taken by almost everyone threatened by patent litigation, is to settle with the patent-holder as quickly as possible for whatever amount of money they demand.

In practice, therefore, an issued patent is a valid patent as long as the patent-holder doesn’t try to sue anyone too large. (And even the largest corporations usually settle.)

That’s the problem.

The patent system is a good idea, in theory.

The patent rules are sensible and should prevent highly damaging patents from being issued, in theory.

The patent office should make every reasonable effort to ensure that they enforce the rules, in theory.

But in practice, this isn’t what happens. It’s not even close.

Good public policy isn’t based on what should be, but what is.

Patents are a good idea. The rules of the patent system were well-designed and have been refined for hundreds of years, mostly for the better. But if they’re not going to be properly enforced, it’s hard to argue that the system is anything but fundamentally broken. And if the rules can’t be properly and consistently enforced, I don’t see how it can be fixed.

We have two software patent articles on the front page of HN. One from a developer saying they are unfixable; another from a lawyer saying they aren’t broken. I think this succinctly describes the situation we are in.

—MatthewPhillips in Hacker News’ comments on my post

Sales of obscure game consoles vs. non-iPad tablets

HP hasn’t released any sales figures for the TouchPad yet. I wonder if it will outsell the Virtual Boy this year.

I didn’t include the iPad’s approximately 30 million units on here because it distorted the graph’s scale too much.

(source, source, source)

Updated iOS device and OS version stats from Instapaper

This is a quick update to my iOS device and OS version stats as reported by Instapaper 3.0.

The same disclaimers apply. This sample does not represent all iOS owners. Specifically, it only represents people who:

So it’s very useful to me, but the usefulness to you will vary.

Here’s the updated data from today:

The biggest differences from five months ago:

Over the last few months, these stats have helped me decide to raise the minimum OS requirement for Instapaper from 3.1 to 4.2 in its next release,1 due out probably within a few weeks. This will break compatibility for the 0.4% of customers still using original iPhones or first-generation iPod Touches.

Development and testing of the new release with iOS 4’s luxuries has been more productive and efficient than ever. My code is cleaner, I’ve deleted a lot of old hacks, and I’m able to do a lot of new features with far less code (like gesture recognizers). If I get an angry email from every one of my 5.4% of customers who don’t (but can) run 4.2, I’ll feel guilty and will probably have a crappy night, but it will have been worthwhile.

  1. Why 4.2 instead of 4.0? Fewer edge-cases, fewer old iPhones needed for testing, and fewer special cases in the code for missing APIs. Every device that can run 4.0 can also run 4.2.

    If I raised the requirement to 4.3, I’d lose iPhone 3G and second-generation iPod Touch owners. If I’m going to cut them off, I’d rather just wait until I can require 5.0, probably in late 2012. ↩︎

Nintendo’s pressure to make iOS games

The story of Nintendo’s investors pressuring them to release iOS games instead of more hardware platforms got a lot of attention this week.

While I don’t think there’s much to it, in that I don’t think this particular event will generate meaningful change in the company’s direction, it raises an interesting question: why should (or shouldn’t) Nintendo stop making hardware and just make a bunch of money selling its popular game franchises on other popular platforms like iOS?

Their hardware profitability woes are relatively new: while the hardcore gamers made fun of the Wii’s awful graphics, shallow gameplay, clunky to nonexistent online abilities, and lack of media-center features, Nintendo was busy selling everyone low-end hardware for $250 that wasn’t particularly fun until you spent hundreds of dollars more on controllers and various game-specific plastic attachments. The Wii kicked ass in the marketplace and made Nintendo a ton of money, so it’s hard to argue that they need to change anything fundamental about their business.

But Nintendo’s greatest strengths — its stubbornness and extreme conservatism — have made it less prepared than anyone to deal with Apple’s nearly accidental takeover of the portable gaming market.

If I were a Nintendo investor… well, I think I’d seriously reconsider that position right now. I’d certainly want to see a big change on the horizon. But I don’t know what that could be.

Nintendo’s financial health depends on a world where they can keep selling many millions of people those plastic steering wheels for $15 and nunchucks for $20 and controllers for $40 and the same games over and over again for $50 on new $250 systems every five years.

I don’t know how much of that world is left, or for how long, but its best times are probably behind us. The game-console world in 2011, especially portables, is a bit like the PDA world in 2004: it’s being eaten by consolidation from more ubiquitous devices. Game consoles will always exist, but they will probably decline significantly in sales over the next decade and be relegated to smaller markets like “hardcore” gamers, much like high-end PC gaming.

Nintendo could certainly sell their same games over and over again for $7.99 in the App Store or try to cheat kids and grandparents out of money on Facebook, but that won’t come close to earning the amount of money they grew accustomed to printing with their hardware business before 2010.

I’m not even sure that their brand recognition is relevant enough in these markets to guarantee that their games would do particularly well there. How much of the iOS gaming market is too young to have much loyalty to Mario and Zelda? (And how much of the Facebook gaming market is too old?)

Nintendo should probably reconsider its direction, especially in portables, but I don’t think this is where it should go.

Rose-colored glasses

Since HP’s announcement yesterday that they’re discontinuing webOS device development (and probably webOS itself, as we know it, but they won’t admit that yet), I’ve read a lot of smart writers eulogizing webOS and mourning its premature death.

HP claims that inferior hardware was the problem, and apparently webOS could run much faster on the iPad hardware. The implication, of course, is that webOS could be awesome and more competitive with iOS in a few more hardware generations, or if HP could license it to someone who could build better hardware. (It’s certainly worth asking why HP couldn’t make adequate hardware themselves.)

The hardware, not the software, was supposedly holding webOS back from being successful.

But looking back at the TouchPad’s reviews, they were almost universally in agreement that the software was just as mediocre as the hardware.


But for every detail that Palm gets right—the re-sizable keyboard that’s pretty nice to type on—it blows something else, like not having a double-tap spacebar shortcut for periods, or the lengthy, complicated mounting process to get your music on there. …

There are critical apps in beta form, like Kindle and some surprises, like Facebook. … It’s funny that the TouchPad is the third tablet platform to run Flash, and it’s also the third tablet on which Flash runs like garbage.

The Messages app was a consistent bag of hurt, refusing to sign on at all sometimes, or to deliver AIM messages, even though I kept receiving them. Email contents wouldn’t show up, often up to 10 seconds after I opened a message. The HP app to get music onto your TouchPad is loathesome—pure HP, and sweet Christ I hope it’s not a sign of things to come for Palm. (Speaking of: Where’s the cloud music?) And there are so many more little problems throughout (ugh, Skype).

Josh Topolsky:

Presses to buttons on the screen would go unanswered, applications would suddenly pause, lists I was scrolling moved intermittently and erratically (or would just disappear altogether). Sometimes the device felt smooth and light, while at other moments it locked up or sputtered to a point of complete aggravation. More than once I had the entire system freeze and then reboot while I was in the midst of navigating (or trying to navigate) my way out of some weird UI fender bender. All across the OS I found myself discovering dark corners of unfinished or untested chunks of the UI, like when I would use the upward swipe gesture to bring up the launcher, and accidentally open an app instead.

And it wasn’t just about speed or smoothness. The new virtual keyboard provided here — while very capable at times — would often not respond or respond slowly to key presses, making for messy emails and messages. …

Skype, while integrated into the core of the OS for calls (more on this below), was generally buggy, frozen, unable to connect, or just downright awkward to use. Rotation on the device was a pain too, as it would often change direction unexpectedly, or alternately fail to respond to orientation shifts when it was asked.

Walt Mossberg:

[On] various occasions, the email app failed to display the contents of messages, the photos app failed to display pictures, and the game “Angry Birds” crashed repeatedly. All of these problems required a reboot of the device to resolve. In addition, I found the TouchPad grew sluggish the more I used it. Again, a reboot was needed to restore normal speed. …

The Web browser generally worked well, but Flash was uneven. Most Flash videos played fine, but some froze or stuttered badly, even on a fast Internet connection. A site written entirely in Flash wouldn’t even load.

Harry McCracken:

In my trials, the TouchPad too often just didn’t work. A few examples:

  • Music playback didn’t always begin the moment I pressed play, and sometimes stammered mid-song;
  • A 1080p movie that looked dynamite on the iPad 2 and Galaxy Tab 10.1 wouldn’t play at all on the TouchPad;
  • The tablet thoughtfully discovered and configured the HP printer that sits on my network. But whenever I tried to print, it told me that my OfficeJet was incompatible.

The TouchPad was at its shakiest when I crammed its RAM with a bountiful supply of apps, Web pages, and other items. Like an overtaxed Windows 98 computer, it would crack under the pressure, rendering screens incompletely, ignoring my input, and (on one occasion) spontaneously rebooting. Closing a few programs restored it to good working order. I know that it’s possible to build a tablet that doesn’t freak out when it’s short on memory, though — because I’ve never seen an iPad or Honeycomb tablet do it.

Third-party hardware won’t fix this.

Some performance problems can be mitigated with faster hardware. But many can’t, as anyone running Photoshop on a Mac Pro knows: fast, multicore processors don’t help most architectural inefficiencies in software.

Even if hardware could someday make webOS consistently responsive, it would likely be working much harder to do it, since so much of webOS is written in high-level “web technologies”. In addition to the performance complaints, most of the reviews agreed that the TouchPad had significantly lower battery life than the iPad 2. That makes sense: the CPU needs to work harder, wasting more power, to manage webOS’ complex interface and high-level architecture.

If that’s the case, it will always be the case. webOS can’t compete strongly if its devices are always significantly slower, have worse battery life, or need bigger batteries than competing devices in the same hardware generation.

Even if performance could somehow be fixed, every review also cited significant problems with bugginess and poor attention to detail. These betray deep-rooted problems in the organization: why was such a buggy, unfinished OS shipped, and why didn’t anyone care about the details?

These seem to be perpetual problems with webOS devices. They’re always supposedly 6 months away from being awesome, but they’re always too sluggish and with too many rough (or sharp) edges.

Anand Lal Shimpi’s criticism of the first Palm Pre in 2009:

Palm needs to work on a lot unfortunately. Synergy needs tweaking, there’s no visual voicemail, limited search functionality, limited copy/paste and there’s absolutely no reason that anything should ever be slower on the Pre than on the iPhone. It’s like me writing software that somehow runs faster on an Athlon 64 than on your Core i7 system. It’s clear that Palm has a lot of optimizing left with the Pre. I’d say there’s a good 6 months of work there to get this thing perfect.

Sounds familiar.

HP definitely mismanaged Palm. The TouchPad’s software shouldn’t have shipped when it did. The hardware wasn’t very good. The marketing was insufficient. The retail channel was poorly managed.

But webOS, despite having some great ideas, never became competitive. Palm and webOS’ developers bear most of the responsibility for that, not just HP’s managers.

This is a high-stakes game. Apple is kicking everyone’s asses so much in the “tablet market” that it’s really not accurate to call it that. Competitors need to be great on day one to stand a chance.

webOS was always “getting better soon”. Maybe we should finally give up our unfounded hope, like HP probably has, that webOS ever could have been great, because it never was.

Steve Jobs resigns as CEO, stays as Chairman

A portion of Steve Jobs’ statement:

I have always said if there ever came a day when I could no longer meet my duties and expectations as Apple’s CEO, I would be the first to let you know. Unfortunately, that day has come.

I hereby resign as CEO of Apple. I would like to serve, if the Board sees fit, as Chairman of the Board, director and Apple employee.

As far as my successor goes, I strongly recommend that we execute our succession plan and name Tim Cook as CEO of Apple.

Apple’s full press release indicates that the board just did exactly that:

Apple’s Board of Directors today announced that Steve Jobs has resigned as Chief Executive Officer, and the Board has named Tim Cook, previously Apple’s Chief Operating Officer, as the company’s new CEO. Jobs has been elected Chairman of the Board and Cook will join the Board, effective immediately.

John Gruber was right about the succession plan (which shouldn’t be a huge surprise). Note that none of Steve’s or Apple PR’s language indicates that Cook’s CEO role is interim or temporary.

I’m sure we’re going to see a lot of opinions and speculation on this over the next few days. I’m not really sure what to say, except that while I think this will cause a lot of short-term doubt about Apple’s future, most of that concern is unwarranted.

Steve has been executing a gradual departure from Apple for the last few years. He’s incredibly forward-thinking, cares deeply about Apple and its future, and has certainly ensured that there’s a plan and a process to completely replace him. This is simply the next step — a big step, granted — in that process.

I think it’s premature to say goodbye. This change effectively makes Steve’s medical leave permanent: while I don’t know the details of his situation, I don’t think it would be unreasonable to guess that he’ll probably be about as involved in the company’s decision-making as he’s been for the last few months.

Regardless, it’s rarely a bad idea to tell someone how much you appreciate their work, even if it’s not yet finished.

Steve, if for some crazy reason you’re reading this:

You’ve defined a generation and changed the world.

Thank you.

We wish you the best.

Top 5 Things Apple Must Change Now That Steve Jobs Isn’t CEO

  1. I don’t know — what they’re doing seems to be working pretty damn well already.

My semi-educated guess on the Amazon tablet

In last week’s Build and Analyze, we speculated about what the future Amazon tablet might be, fitting the most credible rumors we’ve heard so far. (This is basically an expanded version of that, so if you listened, parts will be repetitive.)

Dan Provost made some great points in response, and I think he’s on target except for the screen technology. I don’t think color e-ink is product-ready yet. Even if it could match the resolution and response time of today’s grayscale e-ink displays, that’s still nowhere near good enough to play video, animate anything, or smoothly scroll a page. I doubt that color (or probably even grayscale) e-ink will ever be fast enough for those roles. So I’m ruling out its use for a rich-media tablet, and I think that’s what the Amazon tablet will be, so I’m betting on an LCD screen.

Here’s what I’m guessing the Amazon tablet will be, and I don’t think we’ve heard any credible rumors that suggest that any of the major points are wrong:


Something similar to the Nook Color: an inexpensive, low-end 7” tablet with a backlit color LCD capacitive touchscreen covered in glass. (Or, in layman’s terms: like the iPad, but slower, smaller, and cheaper.)

Amazon will go to great lengths to get very good battery life, even at the expense of thinness or lightness. While it won’t match the e-ink Kindles, it will probably provide at least 12 hours of reading time at low brightness.

Like the Kindle 3, it’s going to feel cheap, but most people won’t care, because it will be cheap.


Android. But not the way you might assume.

The Nook Color runs Android, but you’d never know it: it’s buried under B&N’s completely custom interface and applications. There’s no Android Market to download apps, no Gmail, no Google Maps, no Google-anything. Saying it “runs Android” is like saying the HP TouchPad “runs Linux”: it’s more of a low-level technicality than any indication of what its customers see and use.1

I see Amazon doing the same thing: Android 2.x under the hood, but an Amazon UI with Amazon apps for all core functionality. And, importantly, the Amazon Appstore — and not Google’s Android Market.

This is almost certain, because in all likelihood, they don’t have a choice. Amazon probably wants to have so much control over the software (which is wise) that they’d never pass Google’s “compatibility” requirements for being an officially branded Android device, so Google won’t let Amazon use their standard (and very closed) app suite.

It’s better for Amazon this way: they can control the entire experience, their Appstore will gain a large captive audience, and they won’t rely on Google for any technical or branding dependencies. Few customers would ever realize that they’re using an Android device.


All of Amazon’s content stores will be available and prominent: ebooks, newspapers, magazines, music, video, and apps.

The content, not games or apps or communication or productivity, will be the focus in the tablet’s marketing and its interface.

Video and news-browsing will be major upselling points from the e-ink Kindle and will likely be emphasized in the interface.

Availability and positioning

Kindle is a very strong and successful brand for Amazon. The name will include “Kindle”, such as Kindle Touch, Kindle Color, or Kindle Tablet.

It will be available within the next year, possibly very soon for the holiday season.

It will not replace the e-ink Kindle, but will be an additional product above it, like a “deluxe” Kindle.

It will cost around $249 unsubsidized or $199 with ads (“Special Offers”).

In other words, it will be very similar to the Nook Color, but with the strength of Amazon’s content stores and retail power behind it, which I think will make it a much bigger success than the Nook Color.

I see it taking over much of the low-cost market: people who buy inexpensive gadgets on Black Friday at Wal-Mart, currently served by a bunch of awful, dirt-cheap tablets that geeks like us have never heard of from companies like Sylvania. (But if Amazon can’t or won’t sell it in big-box stores, that’s going to inhibit this role.)

Relevance to the iPad

With aggressive pricing and heavy promotion on Amazon’s site, they’re going to sell a lot of these. It’s certainly going to be inconvenient for the iPad, because some buyers will unquestionably pick the much cheaper Kindle tablet instead. It will probably be the only non-iPad tablet that will sell in enough volume anytime soon that anyone needs to think about it.

But I’m not sure that it will significantly hurt iPad sales. It’s probably not targeting the same market. If people want an iPad, they typically want an iPad — they don’t first decide that they want “a tablet” and then go find the tablet that they like best.

And the iPad’s entry price won’t always be $499. It wouldn’t surprise me to see a cut to $399 for the holidays, and I bet the next iPad’s entry-level price will be $399. But I don’t think Apple plans to (or wants to) compete in the $200-tablet market for the foreseeable future.

The Amazon tablet will be much more like the Kindle: an alternative, inexpensive device dedicated to Amazon’s content stores, bought and used primarily for reading books, browsing the newspaper, and watching movies.

Some people who would have bought an iPad will buy the Amazon tablet instead, but many people will buy it in addition to an iPad, and many of its exclusive buyers wouldn’t (or couldn’t) have bought an iPad instead.

Nothing’s getting “killed”, except maybe some of those cheap Sylvania-etc. tablets, but it’s going to be an important and successful product that will deserve our attention.

  1. Nook Colors can be rooted to run a more typical Android environment, but most customers will never do this. ↩︎

What does the Amazon tablet need to do?

Ben Brooks responded to my Amazon tablet speculation, suggesting that a product like the one I describe would sell very poorly. I disagree.

One important part of his argument is that Amazon would probably design the core productivity apps badly:

The core OS apps are the apps that should be provided on any serious tablet from day one. Those apps include (at a minimum):

  • Web browser
  • Email client
  • Calendar
  • Maps
  • Music/media player
  • App Store, or some way of getting more apps.

I agree that Amazon would probably design these poorly. But I don’t think these priorities reflect actual usage of iPads today or the future theoretical Amazon tablet.1

I see normal people using iPads all the time, and I hardly ever see them using Safari, Calendar, Maps, or Music. Anecdotally, the list consistently looks like this:

This is a very different list, and the media apps can get away with very little UI chrome. This is how the e-ink Kindle gets away with relatively poor interface design: most of the time, you’re seeing almost none of it.

I agree with Dan Provost that a web browser isn’t even necessary.

If Amazon can deliver a $249 tablet that does a serviceable job for reading books, browsing some top newspapers and magazines, watching movies and TV shows, and playing some casual games, that’s going to be very attractive to a lot of people.

  1. If so, we’d see a lot more people trying to use the Kindle’s web browser (which, with the Kindle 3, does use WebKit but still sucks). ↩︎

The romantic notion is the unknown garage startup, the Apple of 1977, but garage startups only succeed in industries that are garage-sized when they start.

Watts Martin