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


The worst cold I’ve had in a decade is beginning its third week. I knew we were screwed as soon as I dropped my kid off at preschool and saw one of his classmates visibly very sick. Fortunately, my kid and wife get over viruses quickly, so they were fine in less than a week.

I started thinking, in my miserable feel-like-I-have-a-fever-but-I-don’t state tonight (during which I probably shouldn’t be blogging, but oh well), what a shame it is that people build up immunities to viruses throughout their entire lives, but when they die, all of that progress is lost, and every new person needs to start all that work from scratch. How incredible would it be if we could somehow capture and recreate those immunities so future generations wouldn’t get these viruses?

It only took a few more seconds before my cold-impaired mind realized that it had just invented vaccines, they already exist, and they’re amazing. Because even though we haven’t found a vaccine for colds yet (and probably won’t), the common cold is mostly a minor inconvenience. Vaccines for much more deadly viruses have existed for decades, most work extremely well with effectively zero risk, and they have saved millions of lives.

Vaccines are truly one of humanity’s greatest and most important accomplishments.

It’s tragic, dangerous, and incredibly destructive that society is needlessly regressing on this front. I’m sadly confident that anti-intellectualism and shunning of widely proven scientific data, selfishly and shamelessly encouraged by entertainers and politicians to advance their careers, will prove to be the most damaging and deadly regression of developed society in my lifetime.

We’ve made vaccination a “personal choice” because anything else is too expensive, politically. Parents refusing vaccines are so numerous now that we can’t protect ourselves from them. We can’t know if the next visit to the doctor or the next day at school will expose us to a dangerous disease that was nearly eradicated a few years ago but is now, tragically and stupidly, on the rise.

A harmless “personal choice” doesn’t damage others. The color of your pants is a personal choice. Vaccine denial isn’t — it’s a severe risk to the public’s health and safety. Not just to the deniers — to everyone.

Refusing vaccines isn’t new, but it used to be very rare and unheard of. Deniers used to vaguely cite religion (another “personal choice” that often damages innocent bystanders in practice), and nobody really questioned it because religious justification is the ultimate conversation-ender, immune to almost any common-sense challenges or legal restrictions.

But now, being anti-vaccine is just another societally acceptable difference of opinion, a pants color, a team you’re on, an option to tick on your Facebook page. The most scary and dangerous thing about anti-vaxxers today isn’t that they exist — they always have, and always will — it’s that their “personal choice” brings almost no consequences or restrictions (unless their children contract a preventable disease, which I wouldn’t wish on anyone), so they’ll only keep getting more numerous.

I don’t know how to fix this. Government-required vaccines are impractical, invasive, politically impossible, and probably wouldn’t actually be effective — there would certainly be a “religious” exemption, so anti-vaxxers would just go back to conjuring vague religious justifications like they used to.

We can’t make people vaccinate themselves and their children. But we can make that a more expensive proposition for them, and a less dangerous proposition for the rest of us, by speaking out and making unnecessary anti-vax choices less societally acceptable.

Even if that means an occasional tense conversation, even if that means my kid can’t go over to certain friends’ houses, and even if that means schools and doctor’s offices need to start turning people away who make that “personal choice”, we need to push back.

Because next time, it might not be a cold.

Google and blogs: “Shit.”

Have a blog? How has your traffic been for the last year or two — say, since July 1, 2013?

Mine’s been clearly flat and slowly declining — the first time the trend has ever gone down — even in periods where I write a lot. I’ve talked to some friends who have seen similar plateaus and declines over the same period, also for the first time. Inbound links from bigger sites also aren’t worth as much as they used to be, suggesting that even big sites are struggling to maintain and grow their traffic.

Nobody’s really talking about it, but I suspect this is a wider trend: blogs aren’t dying, but they are significantly declining. 2015 might be a rough year.

In Is Google making the web stupid?, Seth Godin suggests that the declining prominence of organic results in Google searches is significantly to blame:

If you want traffic, Google’s arc makes clear to publishers, you’re going to have to pay for it.

Which is their right, of course, but that means that the ad tactics on every other site have to get ever more aggressive, because search traffic is harder to earn with good content. And even more germane to my headline, it means that content publishers are moving toward social and viral traffic, because they can no longer count on search to work for them. It’s this addiction to social that makes the web dumber. If you want tonnage, lower your standards.

(Don’t miss the cited Aaron Wall article as well.)

Shallow social-shareable listicles and clickbait headlines have always been plentiful on the web, but it does seem clear that they’re getting much worse and more dominant recently.

Google is making the problem worse, but they’re not the root problem. In fact, the real problem is a pretty big problem for Google, too:

Everyone’s spending increasingly more consumption time dicking around in apps and snacking on bite-sized social content instead of browsing websites and searching Google.

Publishers are relying more on social traffic not because Google’s squeezing them out, but because that’s where everyone went. The dominance of mobile usage, social networks, and YouTube, plus attention-competition from apps, are the real problems for web publishers and blog writers.

The social and app revolutions haven’t been purely additive — much of the time people spend on those now has come at the expense of search, RSS, and bookmarks.

Every hour we spend on Twitter or Facebook instead of reading and writing elsewhere is just making this worse — and I’m as guilty as anyone.

Social networks have powerful benefits and are here to stay. But like any trend, we’ve swung too far in that direction for our own good, as both producers and consumers. I hope the pendulum starts to swing back soon, because it hasn’t yet. It’s going to get worse before it gets better, if it ever does.

If we want it to get better, we need to start pushing back against the trend, modernizing blogs, and building what we want to come next.

Listen to podcasts at whatever speed you want

John Lagomarsino at The Verge, in Stop listening to podcasts at 1.5×, presents an argument commonly made by radio and storytelling purists against speeding up your podcasts (including my Smart Speed feature). Unsurprisingly, I disagree.

I’m a purist for a lot of things. As I’m writing this, I’m drinking Kenya AA coffee, black (of course), that I roasted myself a few days ago. And I’m listening to music — a full album, in order — on what I can confidently say are the best headphones in the world.

But if everyone had to drink coffee black, there would be far fewer coffee drinkers, and the business would dramatically shrink. I also wouldn’t expect much success in the music business if I spent a lot of time pleading with people to please stop listening to shuffled, streamed singles on earbuds and Beats because that’s “not what the artists intended”.

Anyone dictating how people can or should consume media only ensures their own rapid irrelevance.

Lagomarsino illustrates his point with the Mike Daisey “Retraction” episode of This American Life, which famously includes extremely long, tense pauses. This episode’s timing and pauses are the content.

Speeding up the Mike Daisey pauses is like skimming an article — you’re missing some of the detail and experience that the author intended. But a lot of articles aren’t interesting enough to be read slowly and completely, and most people don’t have time for that. People naturally skim and vary their reading speed as needed for the situation they’re in, how much they care, and how much attention they think the content deserves, and many people are simply faster readers than others.

Podcasts (and video) are impossible to skim effectively, but we can vary our listening speed. Just as not every article is worth reading slowly and completely, not every podcast is This American Life. Even most episodes of This American Life aren’t as timing-sensitive as the Mike Daisey retraction. Some podcasts are painstakingly crafted, artistic “storytelling” shows, but most aren’t, by far.

Most podcasts are produced by hobbyists and audio amateurs, effectively none of whom are editing for hours to craft precisely paced stories, and that’s fine, because the best thing about podcasting is that not everything is a big production. Not everything is a professional radio show. Instead, we have a rich breadth of voices and niche subjects, free and on demand, that makes the medium infinitely better than broadcast radio could ever be.

Enjoying the full experience of all media and preserving “what the artist intends” is a romantic ideal, but it’s both overrated and unrealistic in reality. Not everything is that good, not everyone cares that much, and not all media produced is perfect and immutable.

The biggest reason people cite for not listening to more podcasts is that they don’t have the time. My goal with Smart Speed was to directly address that: to make more time for people. And it has: since Smart Speed time-saved totals are synced to Overcast’s servers, I can happily report that Smart Speed has cumulatively saved 55 years of listening time so far. I bet that the vast majority of that time saved was subsequently filled with… more podcasts.

And it doesn’t hurt anyone. I don’t want sugar in my coffee, but it won’t impact my enjoyment of my coffee if you put sugar in yours. I can enjoy the crap out of listening to live Phish shows with my HE-6 regardless of whether you listen to pop music with Beats. And your enjoyment of podcasts at 1× isn’t affected at all by me listening at 1.125× plus Smart Speed. There’s no downside to giving people these options.

If the option to speed up podcasts lets people listen to more podcasts, everyone wins.

The Matias Ergo Pro Keyboard

Matias is known for their mechanical keyboards, and — unusually for mechanical or ergonomic keyboards — available Mac key layouts and function keys. I’ve been looking forward to their first ergonomic keyboard for a while, and it’s finally here: the $200 Ergo Pro.

Ergonomic keyboards are always compromises. Usually, you need to choose between annoying non-standard key layouts, miserably mushy keyswitches, or ugly, clunky designs — sometimes all three. I’ve needed to use them to keep RSI at bay for the last decade, and I’ve never been completely satisfied with any of them. The closest I’ve come to keyboard satisfaction is the $60 Microsoft Sculpt Ergonomic, which is very good overall, but has slightly mushy keys and a few non-standard key-layout annoyances (of course).

The Matias Ergo Pro strikes a similar balance. It’s very good overall, but with a few drawbacks.


Unlike the Sculpt’s one-piece design, the Ergo Pro’s two halves are physically separate and connected by a cable. This is a mixed bag: it provides flexibility, but it’s also frustrating to have no way to lock in your preferred setting, leaving you to figure it out again whenever it’s moved.

Fortunately, the split-halves design on the Ergo Pro stays in place much more than the frustratingly flimsy Kinesis Freestyle 2 because the Ergo Pro is much heavier with more stable feet. If you’re going to have disconnected halves, this is the way to do it.

It’s wired, so it doesn’t make for as clean of a desk as wireless models, but it’s also much more reliable. I’ve had occasional flakiness and reception issues with the wireless Sculpt, and ultimately I prefer the reliability of a wired keyboard if given the choice.

My trick for a clean desk with a wired keyboard: tuck the wire under the keyboard, around the front lip of the desk, and clip or tape it to the underside of the desk running straight back.

Unusually, the USB cable is standard and removable, and the Ergo Pro even comes with two different lengths. And since it’s wired, it can conveniently offer three extra USB 2.0 ports right on the keyboard — a rarity these days — although one of the three ports is puzzlingly located, facing directly into where the other half of the keyboard is likely to be.

The two halves of the keyboard are connected by a removable cable on a self-winding spool with 4-conductor TRRS male plugs on each end — the same connectors at the end of headphones with inline smartphone remotes. This is an unusual cable, but also cheaply and easily replaceable online, which is a very nice touch. But the supplied TRRS cable and spool are bulky and ugly — I’ve ordered one of many mysteriously cheap alternatives with right-angle plugs to tidy up the look a bit.

The main cable, left, with a right-angle Micro-USB plug. The halves connect with a spooled headphone-like cable, center. Above it is the USB port bizarrely facing into the center gap.


Each half of the Ergo Pro has three flip-down legs that can be used in combinations to create either flat, negative tilt, or tented alignments. Negative tilt means the keyboard slopes downward from front to back, rather than the much more common inverse. Tenting means the middle of the keyboard is elevated higher than the left and right sides (like a tent). Both are widely regarded as ergonomically healthier keyboard shapes.

The negative tilt is pretty subtle, and it feels weird without tenting. I didn’t find negative-tilt mode comfortable, so I switched to tenting mode pretty quickly.

You can’t use tenting and negative tilt together, which is unfortunate — neither mode alone is as ergonomically comfortable as the Sculpt for me, but a combination of both might have been.

Top to bottom: Kinesis Freestyle 2, Matias Ergo Pro, Microsoft Sculpt Ergonomic, Hops.

Left to right: Kinesis Freestyle 2, Matias Ergo Pro, Microsoft Sculpt Ergonomic. Top is with the Ergo Pro in negative-tilt mode, bottom is tented mode. (The Kinesis is also capable of many other angles.)

As you can see, the Ergo Pro is noticeably thicker than the others, and the keys are noticeably taller off the desk than the Sculpt’s. To use it comfortably, I needed to lower my desk by about an inch. In addition to the inability to tent and negative-tilt simultaneously, the thickness seems to be the main limiting factor for its ergonomics. It’s still better than traditional keyboards, but not as comfortable for me as the others.

The built-in gel wrist pads are great — by far the thickest and most comfortable of the group. You’re not supposed to actually rest your wrists on them, but nearly everyone does (at least sometimes). I’ll often tend to lean forward and put an elbow on one while thinking or looking closely at the screen, and the Matias is the only one that can support that reliably: the Kinesis can’t support it at all, and the Sculpt’s stand pops off if you put too much pressure on it.

Keys and Layout

Matias is known for using great mechanical keyswitches, and the Ergo Pro lives up to that reputation. These switches feel substantially better than the Sculpt’s laptop-style scissor keys or the Freestyle 2’s light-touch membrane keys.

Compared to most other mechanical keyswitches, these are quieter and moderately springy. They’re still noticeably louder than laptop and membrane keys, but not as loud as common Cherry or old buckling-spring keys. Matias’ sound recordings are accurate: it’s still generally a loud keyboard, but not ridiculously so. If mechanical keyboards usually make your spouse or officemates want to kill you, a Matias keyboard might earn you a few more weeks to live.

The key layout is very good, but not perfect. I wish the Escape key was in the normal position above the tilde and Fn was below Shift, laptop-style. And I’d gladly do without the action buttons on the left, the Home/End cluster on the right, and the Num Lock that turns the right half into a number pad (which I keep hitting accidentally).

But these are really minor nitpicks. Given the inexplicable quirks of other ergonomic keyboards — the Freestyle 2’s non-standard Mac function keys, the Sculpt’s Esc/function chiclet buttons and Fn slider — the Ergo Pro is refreshingly normal. It’s mostly a native Mac keyboard with standard Mac function keys, and all of the functions work without installing any software.

You don’t want your keyboard layout to be exciting and different, and this isn’t, thankfully. It’s a high-quality Mac keyboard, cut in half. That’s exactly what a split-ergonomic Mac keyboard should be.

I Like It

The Ergo Pro is very good, but so is the Sculpt Ergonomic. I thought I’d have a harder time making a recommendation between them, but now having used both, the difference is clear.

The Microsoft Sculpt Ergonomic is a more comfortable ergonomic keyboard, but the Matias Ergo Pro is a better overall keyboard. The Ergo Pro has better-feeling keys, a more reliable wired connection, better function keys for Mac, and USB ports.

If you need the highest degree of ergonomic comfort, the Sculpt is probably the better choice, as long as the one position it offers is compatible with your ergonomic needs. (It’s also significantly cheaper.)

If you want a high-quality mechanical keyboard that’s split down the middle and ergonomically better than most standard keyboards, check out the Ergo Pro.

The first run is sold out, but you can preorder for expected delivery in April.

Thanks to Matias for supplying an Ergo Pro for review.