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

Lasting value

I recently had a good reason to look through my blog archive for a handful of articles that were very good, relatively timeless, interesting to a broad audience, and G-rated that didn’t include any references to celebrities, political figures, or well-known trademarks. I thought I could easily come up with 5–10 suitable picks.

I found very few. And the most recent pick, the weakest by far, was almost a year old.

It was sobering.

Most of my favorite writing over the last few years was about specific products or technology companies. There’s a place for that, but in one year or three years or thirty years, who’s really going to care about the politics of technology and the nuances of gadgets in 2012?

My primary outputs, professionally, are software and writing. This is what I’m contributing to the world. None of the software I write today is likely to still be in use in thirty years, but if I write a truly great and timeless article, that could be valuable to people for much longer.

I’m going to continue to write about what’s happening in our industry. But I’m also glad that I had this chance to step back and get some perspective on my work, because I haven’t written nearly enough articles recently that I’ll be proud to show off more than a few months from now.

Streamlining U.S. currency

Canada recently decided to kill its penny. The U.S. should take this idea even further: let’s eliminate the dime, half-dollar, $1 bill, $2 bill, $10 bill, and $50 bill.

People generally seem OK with denomination steps in which the larger denomination represents up to five of the next-smallest denomination. We currently have:

By making my proposed eliminations, we’d have a much more streamlined system and no denominational jumps greater than 5x:

Much simpler.

Bathroom fan timer switches

Bathroom fans are great. Half-baths have one great reason to have one, and bathrooms with showers have two.

First of all, if you’re in the market for a bathroom fan — granted, not a frequently purchased item for most — you should definitely consider the Panasonic WhisperCeiling. It’s much quieter than any other household bathroom fan I’ve ever heard. I got an 80 CFM for a half-bath and 110 CFM for a full-bath, although I think the 80 would have been sufficient for both and it’s significantly quieter than the 110.

The problem with bathroom fans is that they’re inconvenient to use: ideally, you want the fan to run for a little while after you leave the bathroom. If you turn it off when you leave, it’s not very effective, and if you leave it running, you’ll probably forget about it for hours and waste tons of electricity. It may even be a fire risk.

A timer switch fixes this problem. After convincing my skeptical wife to let me install them, she only had two criteria: they couldn’t be the old giant-knob style, and they had to look good.

Amazon research led me to two top models: the fancier-looking Lutron Maestro MA-T51 and the more utilitarian-looking Leviton LTB60-1LZ. We couldn’t decide between them, so I got one of each:

From left: Lutron, Leviton.

They both work. (Exciting.) But beyond that, they’re very different.

The Lutron looks nicer, but in use, it’s more complicated: the small rocker on the right side selects the duration from the numbers on the left, and the giant center button is the on/off switch. The currently-selected time LED glows constantly, even when off. When on, the LEDs pulse down sequentially from the selected time to the bottom in a cycling animation, starting lower as the remaining time nears zero. It’s visually and conceptually busy and overly complicated.

The Leviton is simpler: just tap the time you want, and its LED glows. If you tapped “30M” but there are only 20 minutes left, only the “20M” LED will glow, and so on. When off, the bottom-center LED glows orange. If you change your mind about the time when it’s on, just tap a different duration. To lock it on, hold down one of the time buttons for a few seconds, and the bottom LED glows green.

Visitors have been confused by both. Many have figured out the Leviton, but none have successfully used the Lutron — even a contractor working on the bathroom was scared to touch the Lutron and asked me how to turn the fan on.

The Lutron is also more work if you often choose different durations. After a shower, you probably want the full hour to clear out the humidity, but after ruining the air in a half-bath, you probably only need 15–20 minutes of runtime. With the Leviton, you just tap the button for whatever duration you want each time. With the Lutron, you’re effectively using a menu to change the duration whenever it’s not the same as the previous selection.

Both switches could be even simpler by offering fewer durations. I bet a switch that just offered 15- and 45-minute durations would be fine for nearly everyone. (The Leviton comes in a few variants with different durations, but always the same number of them. The 30-minute-max model is fine for a half-bath, while the 60-minute-max model is probably better for showers.)

The difference between them isn’t big enough for me to replace either with the other model, but the Leviton is definitely my favorite. My formerly skeptical wife loves it, and is now convinced of the value of bathroom fan timers.


(And if you aren’t comfortable doing moderately difficult switch wiring, hire an electrician to install it. It’s not as easy as a regular light switch.)

Why I write about bathroom fans and pillowcasing strategies

Last week, I wrote that I was disappointed to find so few of my recent posts with long-term, general relevance, and I was going to make an effort to increase the number of such posts in the future. So when I posted about bathroom fans and timer switches a few hours ago, I got a few responses that were mocking me for writing something seemingly frivolous after making such a proclamation.

To clarify, I only said I was going to increase the number of long-term-value posts, not that I was never going to post anything light or fun in the meantime. And I think people who complain about my seemingly frivolous subject matter probably have a different idea of “lasting value” than my interpretation.

Installing timer switches on my bathroom fans has eliminated a daily annoyance from my life. Deciding on a pillowcasing strategy has made me sleep more comfortably every night since I wrote that stupid article seven years ago. By caring about what type of headphones I use, I satisfy my own needs better and annoy the people around me less every day.

These may sound trivial, but they add up. As Joel Spolsky wrote 12 years ago:

So that’s what days were like. A bunch of tiny frustrations, and a bunch of tiny successes. But they added up. Even something which seems like a tiny, inconsequential frustration affects your mood. Your emotions don’t seem to care about the magnitude of the event, only the quality.

And I started to learn that the days when I was happiest were the days with lots of small successes and few small frustrations.

I’m constantly seeking other little ways to improve my life: ways to eliminate frustrations and create even more tiny successes and little delights every day. As I wrote four years ago:

I try to be discerning in everything, because I love it. I love the research and acquisition of specialty things, I love finding new and better versions of the things I like, and I love discovering the immense depth of hobbies and goods that most people never see.

All of these minutiae — every little thing I care about in my life, some of which I take the time to write about — adds up to a life full of little victories, and I’m extremely happy.

By sharing insights and opinions on these little things, I can make other people’s lives better, too, a little at a time.

When I first started writing, I was reaching dozens of people. Now I’m reaching hundreds of thousands, a great honor. But the far more satisfying honor is when I hear from a few of them after writing one of my life-minutiae posts, and they tell me that I’ve just made their lives a little bit better.

That’s why I do this.

If you don’t care about such minutiae, that’s fine. I just hope you have something that you do care about. But I care about this sort of thing, and I get immense joy and satisfaction from improving the minutiae in other people’s lives.

Bose SoundLink Bluetooth speaker review

I use a Jambox in my bathroom during the same two situations that necessitate a fan timer. When showering, I like to put the Jambox on a shelf just outside of the shower and listen to podcasts, but it’s hard to get enough volume out of the Jambox for speech to be clearly audible over the shower and fan noise.

The about-to-be-released Big Jambox caught my attention to solve this problem, but reviews convinced me to order the Bose SoundLink Wireless Mobile Speaker instead.

The SoundLink looks decent and has great-feeling build quality, as you’d expect from Bose.

It produces a very impressive amount of volume relative to its very small size, as you’d expect from Bose.

But it also produces a very strange tonal balance: the mid-highs are weak, and surprisingly for something this size and battery-powered, the bass is ridiculously strong and boomy. It’s completely imbalanced.

As you’d expect from Bose.

I played a few different kinds of songs, and in all of them, the boomy bass was distractingly strong. It’s all I could hear.

And that was just in my office, a normal room. When I tried listening in the bathroom, a much smaller room, the bass completely took over like a horrible aftermarket car stereo. It was unlistenable.

The original Jambox, not known for particularly great sound quality but very good for volume in a small package, is much more listenable.

Maybe it’s just my preferred tonal balance: neutral. I leave EQs off and I listen with close-to-reference headphones and bookshelf speakers. I don’t like standalone subwoofers and don’t own any. But this is an unusual preference: most buyers like big, boomy bass, and Bose is clearly designing for them. That’s good for business, but not good for me.

I’m returning the SoundLink to Amazon, and I just ordered a Big Jambox. Hopefully that will be better.

The rumored thin 15” MacBook Pro with Retina, USB 3

9to5 Mac posted a rumor about the next 15” MacBook Pro being in a much thinner (but not wedge-shaped) case, having a Retina display of an unspecified resolution, and having USB 3 ports.

It sounds plausible. Assuming that’s all or mostly true:

Thin case

The optical drive’s gone, but at this point, that’s neither radical nor newsworthy. Its removal saves a lot of space, which is nice, but it won’t save much weight — the optical drive is extremely lightweight.

To achieve the thinner case and reduce the weight, I’m curious to see if they finally removed the glass in front of the screen. On the current 15” design, the glass adds around 0.4 pounds over the matte option, and its extremely high reflectivity is problematic for a lot of people. If the new 15” offers a plastic-glossy screen instead, like the MacBook Air’s screen, that would save a lot of weight and be far less reflective for people (myself included) who don’t like glass screens.

The 13” MacBook Air is 3 pounds and the current 15” MacBook Pro is 5.6. Assuming Apple drops the optical drive and glass screen, they continue to offer at least the same battery life, and they continue to use 45W CPUs and discrete GPUs, I’d expect the new model to weigh 4.5 to 5 pounds. If they drop the GPU and pursue lower-wattage CPUs, they might get a more significant reduction, and the reduced power demands and heat output would result in a better computer for most use.

The thin case also means that Gigabit Ethernet and Firewire 800 won’t fit and are probably just being dropped. Expect video pros to complain.


I suspect USB 3 is going to do to Thunderbolt what USB 2 did to Firewire 800: serve as the dominant interconnect for most peripherals, with the more-expensive Thunderbolt being relegated only to high-end niches.

Thunderbolt has been out for over a year, but there’s still a disappointing lack of peripherals. The few that are available are very expensive, and many potentially useful ones — such as Gigabit Ethernet or Firewire 800 adapters — don’t exist yet as standalone peripherals. (You can get a Thunderbolt-to-Gigabit-and-FW800 adapter for $1000 with a free Apple Cinema Display attached.)

It will be interesting to see if Apple addresses this rumored MacBook Pro’s lack of Firewire 800 and Gigabit Ethernet by making adapters available, and if so, whether those adapters use USB 3 or Thunderbolt.1

Retina Display

I’m not sure I’d want a Retina MacBook Pro yet. I suspect that adoption of Retina assets among Mac apps will be slower than we saw with Retina iOS devices, and more importantly, Retina graphics for websites will likely take significantly longer.

Since non-Retina graphics look worse on Retina screens than on older screens, Retina MacBook users would have significantly worse-looking web browsing for a while — probably years, not months. So I don’t think I’d rush out to get a Retina Mac, but I wouldn’t necessarily avoid a Retina screen when it comes time to upgrade for other reasons.

  1. Apple has sold a 10/100 USB Ethernet Adapter for years, but since it’s limited to USB 2, it’s extremely slow. A modern MacBook Air can generally transfer files faster over wireless (if it’s 802.11n) than over the USB Ethernet Adapter. ↩︎

Speculation on the next Mac Pro

The Mac Pro seems like a ridiculous computer in the age of quad-core laptops. Even I thought so when I switched to a MacBook Pro last fall. But I had some major issues that have only been partially resolved. And in many other ways that I didn’t expect, this setup is far more complex, less elegant, and less reliable than a Mac Pro.

I really hope Apple isn’t done with the Mac Pro yet, because a Xeon E3/E5 update would be awesome.

Back in March, I speculated that the dual-2.9 GHz E5s would reach around 36,000 on Geekbench (64-bit). This now looks like it’ll be correct: Rob-ART Morgan just tested a PC workstation with a pair of the new, Mac Pro-ready Xeon E5 CPUs, with 16 total cores at 3.1 GHz. It scored a whopping 41,242 on Geekbench in Windows.

The fastest Mac Pro you can buy today, the two-year-old, $6200, dual-X5670 2.93 GHz model, scores “only” about 24,262.1

I’m still not sure that the Mac Pro delay is a sign of Apple not caring about it. Remember, they haven’t been waiting for two years because they didn’t feel like updating the line: it has taken Intel two years to deliver the next generation of Xeons.

Assuming Apple still cares a bit about the Mac Pro, what might an update be like?

I don’t think we’ll see a switch to consumer-level CPUs or the price drop that could bring, since consumer chips would require Mac Pro buyers to give up a lot of what they need. If the Mac Pro is going to continue to exist, it should still be a Xeon workstation.

The current Mac Pro is so old that it’s the only Mac still shipping without Thunderbolt, since it predates the introduction of Thunderbolt by six months. Obviously, Thunderbolt would be nice to have on the new Mac Pro. But so would USB 3.0, which inconveniently isn’t built into Intel’s Xeon-compatible chipsets yet. (It’s supposedly coming to the Ivy Bridge updates to the laptops because Intel’s Ivy Bridge chipsets natively support USB 3 for “free”.) Incorporating either of them will require dedicated controller chips on the motherboard. This doesn’t preclude them, but makes them slightly less likely if Apple isn’t particularly dedicated to getting them in there.

I’d love to see the optical-bay area and SATA controller modified to natively support a cluster of four 2.5” drives to accommodate SSDs. (Current Mac Pros can be modified with aftermarket bay adapters and SATA controller cards, but that’s a complex and messy solution.) To make room without other case modifications, one or both optical bays could be removed, or the optical drive could be changed to a slot-loading, slim laptop drive.

I suspect that Blu-ray will still be ignored, mostly for software reasons: while you can buy any BD-RW drive on the market and read or burn data discs from a Mac, you still can’t legally watch Blu-ray movies on OS X. Advertising Blu-ray support without being able to play Blu-ray movies is questionable, and it serves Apple’s strategic interests better to continue pretending that Blu-ray doesn’t exist. Anyone who wants to add a BD-RW drive to their Mac Pro, internally or externally, can do it themselves very inexpensively.

One big question is Retina-display support. If Retina displays are coming to the MacBook Pro soon, would they spread to the rest of the Mac line within a year or two? If so, wouldn’t the next Mac Pro generation need to support them, presumably with the release of a Retina Thunderbolt Display? That’s a big requirement alone, and it would also require special video cards that could drive two or three of them. (The sheer amount of manufactured pixels and GPU throughput required to pull that off makes me think that Retina Thunderbolt Displays might not exist for a while.)

The lack of a Mac Pro update until this point is slightly suspicious, but not cause for much concern. I’m guessing that Apple’s holding back the Mac Pro until USB 3 debuts in the MacBook Pro, a Retina Thunderbolt Display is available, or Mountain Lion is released. If we don’t have a new Mac Pro by the end of the summer, I’ll start to be worried for its future.

  1. To put Mac Pro performance into perspective, the fastest Mac outside of the Mac Pro line is currently the 3.4 GHz 27” iMac at about 12,532, with the recent 2.5 GHz i7 MacBook Pro close by at 11,851. That’s approximately the same performance as a midrange early 2008 Mac Pro. ↩︎

Review: Tonx Coffee

People always ask me how they can make great coffee.

I’ve never had a good universal answer. The real answer is to get an expensive grinder1 and an inexpensive AeroPress, then brew freshly roasted beans and drink it black.

Anyone with enough money can buy the equipment, but most people don’t have a good source of freshly roasted beans. Very few people live within a convenient distance of a coffee roaster, and even for those who do, the nearest roaster might not be particularly good: it might roast poor-quality beans, or it might roast them much too light or dark for your taste.

I solved this problem by home-roasting. But home-roasting is impractical, time-consuming, and fussy. If you’re asking yourself whether you should home-roast, the answer is definitely “No.” (I shouldn’t, either, but I do it anyway.)2

I wasn’t able to roast for a few months during a recent home renovation, so I signed up for Tonx, a coffee subscription service. The deal is simple: every two weeks, they mail a 12-ounce bag of freshly roasted coffee to you for $19. (It’s billed at $38 every 4 weeks, so it’s effectively $19 per bag.)

As coffee goes, this is very expensive. But for high-quality, freshly roasted, mail-ordered coffee, it’s competitive: ordering a similar 12-ounce bag from Stumptown runs about $23 after shipping.

No good mail-order service will compare well to buying it locally: Tonx is effectively $25.33 per pound, while I’ve rarely seen beans at local roasters priced above $16 per pound. But even “expensive” coffee isn’t completely out of reach: if you use 15 grams of beans per cup, a cup brewed with Tonx beans only costs about $0.84. By comparison, $16-per-pound local coffee is about $0.53 per 15-gram cup. We’re not talking about a lot of money either way.

Since I signed up on February 1, Tonx has been very consistent: they’ve sent very good coffees reliably every two weeks, and I usually get them about 3 days after they’ve been roasted, since they’re sent via Priority Mail from Los Angeles.

This is close to ideal. Coffee actually doesn’t taste very good right after it’s been roasted — the full flavor takes 2–3 days to develop.3

While every Tonx coffee has been very good so far, none of their picks have blown me away. This might be because they roast a bit too light for my taste:

Left to right: A local roaster, Tonx, my home-roast. Tap to enlarge.

I’m not qualified to tell you exactly what roast level they use, but my home-roast above is between City and Full City.4 Maybe their roasts are a fairly light City. Regardless, I like it a bit darker, but I think most coffee nerds would be very pleased with their roasts.

Their customer service is excellent: my first shipment got misrouted by the USPS and was going to be almost a week late, so they sent me another one, expedited, at no additional cost. That was the only issue I’ve had to date, and they handled it extremely well.

Tonx is a great option if you want great coffee delivered to your house without having to think about it.

Now, I have a great universal answer whenever anyone asks me how to make great coffee: Get a burr grinder, get an AeroPress, and subscribe to Tonx.5


  1. A lot of people have asked me about manual hand-crank burr grinders. I have the Hario MSS-1B for occasional travel use, and it’s acceptable for a coarse or medium grind (although people nearby will make fun of you). But it takes far too long and far too much cranking to achieve a fine grind, and since the AeroPress is best with a fine grind, I don’t recommend using a manual grinder with it. ↩︎

  2. If you decide to home-roast, well… you really shouldn’t. But if you decide to do it anyway, get the Behmor. Roast under a window and put a good box fan on the sill, blowing out. Open a window on the other side of the room for intake. ↩︎

  3. I once imagined inventing a truly all-in-one coffee machine that would roast, grind, and brew the beans freshly for each cup. The flavor-development delay after roasting is one of many reasons why such a machine should never exist. I have many terrible ideas. ↩︎

  4. Normally I roast to Full City, sometimes just before Vienna, but that would be a poor choice for a fine Kona. ↩︎

  5. You can even save some money on a great grinder if you buy it via Tonx after you subscribe. See the “Tonx Perks” section in your account’s control panel. They ask that people keep this relatively quiet, so it’s in a footnote. Nobody reads footnotes. ↩︎