Ars Technica recently published a good overview of air traffic control technology. A few weeks back I got to visit Norcal TRACON, the central hub that controls the airspace around the Bay Area, Sacramento, soon all the way to Reno. The visit was really cool, I left with a real respect for the skills of the controllers. Also impressed by the technology. Relatively simple but highly customized user interface for the main radar scope display with lots of good human factors design. And similarly simple-but-customized auxiliary information displays, communications, etc. It all seemed very functional.

When I was first learning to fly the controllers intimidated the heck out of me. They were literally the Voice of Authority, they were speaking a special language, they were interrupting my concentration, and I was afraid I'd say something dumb and embarass myself. I'm entirely over that now. A key experience was at Palo Alto with a trainee controller in the tower. I was practicing landings and the tower kept screwing up, at one point even clearing me to taxi on to a runway someone was landing on! No danger: the supervisor immediately corrected the clearance and I wasn't going to taxi into the airplane I could see, anyway. But it was all a good lesson that the controllers are people too and that they're there to help me, not enforce regulations. You also get to know voices and they're quite friendly and helpful when able. It's a good relationship.

ATC has been in the news lately with the story of a controller letting his kids give instructions on the radio. It's been way blown out of proportion, I can't imagine there was any conceivable threat to safety. But there's appearances to maintain and apparently will be hell to pay for a father showing kids his pride in his work. It's a shame: one of the things the TRACON guys told me was they're having a very hard time training new controllers. All the folks hired in 1981 after Reagan busted the union are reaching mandatory retirement age. They're trying to recruit more controllers and accelerate training, but it's not working out as well as they need. After this mess the ATC job must just look all the less appealing. I have absolute trust in the controllers who have helped me, they deserve a little slack and the ability to exercise their own judgment.

I have a love/hate relationship with Apple products, particularly my beloved iPhone. Today Apple proudly announced they're suing HTC over phone patents: that's firmly on the hate side.

We can sit by and watch competitors steal our patented inventions, or we can do something about it. We've decided to do something about it. We think competition is healthy, but competitors should create their own original technology, not steal ours.

— Steve Jobs

Everyone in the tech industry knows the patent system is entirely broken. Other than a few sleazy actors like Intellectual Ventures there's an understanding in the innovative side of the tech business that you don't file aggressive patent lawsuits. You write a lot of patents, you file defensive lawsuits and countersuits, but in general you don't use your patent portfolio as a big club to try to destroy competitors. Apple's taking a big crap on that detente. It's pretty hard to love a company that is going to exploit the broken patent system to stifle innovation. Not that this is new misbehavior, see also Apple's 1988 look and feel lawsuits.

What's particularly galling is Jobs' language about "stealing". We'll know more when the patents are named, but I have to wonder. Does he think anyone else using a multitouch UI is stealing? Pinch-to-zoom? Seamless wifi/cellular networking? I wonder if Jobs called it "stealing" when Apple took the mouse and windowed GUI metaphors from Xerox?

One thing I love about text chat in instant messengers is I can carry on two simultaneous conversation threads with a person at the same time. I write something on topic A, and while waiting for my friend to respond I say something about topic B. When he's done responding to topic A and waiting for me, he can start talking about topic B. Back and forth, ABABAB. Two conversations interwoven, very natural.

Flying airplanes is hard, but pilots do learn to fly planes safely. How do we manage all the complexity? By managing our attention, by using checklists, and by taking advantage of simplifying technologies. Some of these techniques are applicable to other complex tasks, like cooking or software engineering.

Attention training. Practice makes seemingly impossible tasks easier. When I first started flying it took all my attention to keep the plane level. I literally could not handle the radio interruptions. Now that basic flying comes naturally I have plenty of spare attention to talk on the radio. Except when I'm very busy; I've learned to ignore the radio if I don't have time for it. Training has developed my attention capacity and taught me to manage distractions.

Checklists. There's some 35 steps I do to get a plane ready to fly. Fortunately I don't have to memorize them all, because I have a checklist that I follow every single time I start a flight. If I follow that checklist I'm fine. If I forget a step the plane won't start, or my gyro compass will be off a few degrees, or the flaps will be set wrong and I'll stall on takeoff and die. I'll be using my checklist every single flight for the rest of my flying career. Checklists are broadly applicable to many areas. I once wrote a two page checklist for a server deploy at Google and it paid off in spades for years afterwards.

Autopilot. I'm not using an autopilot during my training, but even the simplest one that does nothing but keep the wings level makes flying a lot easier. Ken's about to get a fancy autopilot that can literally fly an entire course for you, including instrument approaches, from about 400 feet after takeoff to 400 feet from landing. Of course you never want to fully trust the robot with the flight, but it's a big help to offload the basic control inputs so you have more attention for more complex things.

GPS navigation. Moving map GPS is a huge user interface improvement for navigation. No more trying to spot landmarks from the air or figure out your position via triangulation. You just punch in your destination, press "direct", and you've got a flight plan. A GPS with a terrain database helps you avoid hitting mountains. Add a synthetic vision system and you don't even need to be able to see out the window. Unfortunately very little modern flight training uses the GPS, it's going to take a generation for common practice to change.

Writing software used to seem very difficult. But over the years I've learned how to code and the technology for writing software has improved. Garbage collection is like autopilot for memory management, modern IDEs are like GPS maps for visualizing code. Unfortunately the pace of technology development in aviation is very slow compared to software engineering because of all the testing and certification required. But there's more opportunities for improvement via training and plenty of checklists.

I'm at the point flying now where I'm fine if things go as planned. But I don't have much experience improvising. Today's solo flight was a good exercise. The weather wasn't as good as predicted so I decided to just do a few trips around the pattern. The reported 240@6kts winds were well within my comfort zone. The wind sock was flopping around a bit, so I asked for an updated wind reading and was told it was still 240@6. OK, that's fine then.

The first complication happened on my downwind. I was following another Cessna around the pattern, cleared to land #2. And following, and following. Normally you turn base at about half a mile or so past the runway. I was more than a mile and a half out still following this guy on downwind when the tower called to remind me I was #2 to land, implicitly asking why I hadn't turned yet. I quickly reported "following traffic still on downwind". That must have woken the other pilot up, he started to turn. I was all out of my usual routine trying to figure out what power and flaps I needed for this greatly extended pattern. Rule #1: fly the plane. I took stock and realized I was way too slow, so I pitched down, added some power and made my turn behind the other plane.

I just about had this unusual approach figured out when I realized the plane I was following was staying high like he was going to overfly the field. Then the second complication: he did a forward slip, dropped swiftly, and landed. And kept rolling down the runway, leisurely using the full length of the field. And me coming in as slow as I dared hoping to land behind him. Tower asked the guy to expedite getting off the runway, he casually replied he'd turn when he could. When I was about 20 feet from touching down the guy was still on the runway so tower told me to go around. Unexpected, but not dangerous.

In the meantime I'm hearing pilots ask for updated winds and now they're 140@9kts. The third complication: the wind had swung around and was getting strong enough to be uncomfortable. Tower mentioned considering switching runways but had five planes in the pattern. Soon they were telling everyone to land and wait, clearing the air. Meantime I'm still up there, seemingly forgotten, and with no idea where the other planes were. So I called in "Tower, am I cleared to turn base?". Our tower guys at KSQL are great. He immediately told me to turn left 180 and enter left downwind for the other runway. So I made my turn, watching very carefully for planes, and proceeded to make a very ordinary approach. Except the ground was going by very fast on short final, the wind had swung around again and I ended up landing with a bit of a tailwind.

That was enough adventure for one day, so I called in to taxi to parking. I'm a bit proud that Tower replied with a "thanks for the help". Guess I got it right!

When I started learning to fly it seemed impossibly difficult. Flying seems easier now, although still challenging. It's roughly the same kind of skill as driving a car and no one thinks back on learning to drive as particularly difficult. But it only takes 6-10 hours of driving time to get a car license, compared to 40-70 hours of flying time for a pilot's license. Why is flying so much harder?

Control complexity. Cars only steer in one dimension; planes steer in two. Even a level turn is hard in a plane, you have to coordinate two controls, except sometimes you deliberately uncoordinate them. Managing engine power is harder in a plane: two or three controls in a piston, not just a single pedal. And then there's auxiliary controls you have to use occasionally: flaps, carburetor heat, fuel tank selector, etc. Even starting a plane requires carefully using four controls in the proper relationship.

Being in the air. Planes at altitude have a lot of potential energy. A primary challenge of landing a plane is learning how to bleed that energy off to make a soft landing. In a car you just brake. And if something goes wrong in your car you can pull over to the shoulder and stop; that doesn't work so well in a plane. So there's hours of learning how to get a plane safely down in various kinds of emergencies and equipment failures. Also learning to ensure you can land: running out of gas or flying in to clouds are dangerous, so there's a lot of training to avoid those situations.

Instruments. Operating a modern car is pretty easy, you don't think about the engine unless an idiot light comes on. Flying requires a lot more vigilance. You're constantly watching your heading, airspeed, altitude, attitude, and engine power to make sure you're flying roughly right. And there's a bunch of extra instruments you need to scan regularly; oil pressure, outside temperature, voltages, etc. It only gets more complicated if you're flying when you can't see, that's a whole 'nother 55+ hours of training.

Communication. Most general aviation flights are constantly on the radio, talking to ATC to get permission to takeoff and land, to get separation from other aircraft, or simply to be followed for safety. We know talking on a cell phone while driving is remarkably distracting. It's not quite as chatty in a plane, but it's necessary communication and it has its own special language. Learning to work the radio was remarkably difficult for me and it's still a constant distraction.

Navigation. In a car you just follow the road and signs, maybe look at a map or a GPS to make sure you make a turn. In a plane you're largely on your own in three dimensions. If it's good weather it's not too hard to avoid mountains, but some of the obstacles are invisible. And while moving map GPS has revolutionized GA navigation, I still have to learn to fly without it: dead reckoning, pilotage (aka "looking down"), flying between radio beacons. And even if you know which way to fly, you still have to calculate for invisible winds.

The good news about pilot training is it seems impossibly difficult at first but gets easier with training. In a followup blog post I'll talk about some of the ways pilots manage complexity. But no matter how skilled the pilot or how many good tools they have, flying always remains a complex activity requiring attention, thoroughness, and constant practice.

I'm doing some geographic hacking. I get confused about degrees, seconds, fractions, and precision of coordinates. Here's some approximations for how far distances are for latitude degrees. Imperial measures, my apologies to my non-American friends who are not crippled by years of non-metric education.

1 degree	70 miles
1 minute	1.15 miles
1 second	100 feet
0.1 degree	7 miles
0.01 degree	3600 feet
0.001 degree	360 feet
0.0001 degree	36 feet
0.00001 degree	4 feet

Those numbers are for latitude. Longitudes get narrower at the poles. At San Francisco's latitude of 37 degrees, a degree of longitude is about 55 miles, or about 80% of a degree of latitude.

A modern consumer GPS with WAAS should be accurate to within about 15 feet. So for many contemporary geocoding applications, you want to record and publish numbers to about 0.1 seconds or 0.00001 degrees. That's 26 bits of precision, which fits comfortably in a fixed-width 32 bit number but is too precise for a 32 bit IEEE float.

One of the fun things about pilot training is learning to do things the old way. I've just learned the art of plotting courses using an E6B, a glorified slide rule. Initially I mocked this old tech, but having learnt it I admire how simple and unmediated it is. And I can do all my calculations on a 65 year old whiz-wheel, my father's Army Air Force issue Computer: Dead Reckoning. The interesting side of the E6B is the wind correction part. If you've got a 20 knot headwind blowing 50° to your left, what angle do you need to fly straight and how much slower will you be going? It's a simple trigonometric problem, but it's too abstract writing it out and doing the arithmetic. On an E6B you just draw the wind vector, right on the computer, then rotate it to visualize the wind vector vs. your path. Then you read the course correction and speed change off the calibrated graph underneath your drawing. It's like a thousand little vector calculations were done at once and drawn on the paper for you to measure against. Very intuitive.

The other side of the E6B is a simple circular slide rule for doing multiplication and division. It has scales for fuel consumption, distance and rate calculations, converting °F to °C, etc. Initially I used my iPhone for this arithmetic but I kind of prefer rotating the wheels and reading the numbers off the scale. The imprecision of a slide rule reinforces that everything is just an estimate. And I'm constantly sanity checking everything to make sure I'm didn't misread a scale by a factor of 10. That extra work seems to help reinforce the course planning.

I'm pretty sure as soon as I pass my pilot exam I'll never use an E6B again. It's much easier to use map-based flight planning software, or for that matter just to jump in the plane and let the GPS computer figure it out for you. More accurate, too. But the old way's kind of cool.

Video totally worth watching, the Vice Guide to Liberia. A travel documentary looking at Liberia and its recovery from brutal civil war. Parts 1-6 are online now, rest to come. The trailer gives you an idea what to expect. Vice is obnoxiously hipster, but you have to respect they actually produced this video. Not something you're likely to see Anderson Cooper do. Liberia is fascinating to me because of its special relationship with the United States. It's essentially an American colony, a republic established by freed slaves in 1847. The civil wars there were horribly brutal and marked by bad action all around, including American and UN supported forces. There's some hope for stability now, but what do you do with a generation of child soldiers raised on drug-fueled killing? Vice's video gives some insight.