The Seattle Times hits the mark on the death of global Internet access: Aerospace reporter Dominic Gates offers the best summary in the media of what the end of Connexion by Boeing truly represents: An end to Internet access in flight anywhere in the world (not everywhere, however). (The second death for Boeing is that it's the end of their plans to build multibillion-dollar satellite-based businesses.)

Boeing built Connexion around a late 1990s business plan, and I had always seen its post-9/11 existence as proof that it was the then-CEO's pet project. There is no love for Phil Condit in Seattle, where I live, as he decided for what appears to be no particular reason nor financial advantage to have moved the company's headquarters to Chicago. Most of its operations and a lot of key staff remain here, while Condit was ousted in 2003 over fallout from the US Air Force tanker deal.

In the late 1990s, jet fuel prices were much lower, and by the time 2001 rolled around, planes were full and airlines made apparently more profit during that short period than in the entire previous history of the industry. Putting 800 pounds of equipment on a plane and charging $30 a flight to use the internets at 35,000 feet seemed completely reasonable. And remember that domestic carriers had signed off on the plan.

Then 9/11 hit. The airline industry sunk. The economy briefly supertanked. Travel seemed less fun and necessary. And then gas prices shot up and up and up. An 800-pound monkey now seemed untenable to airlines heading into and then fully in bankruptcy.

By the time the next CEO came in, well, he was tied up with the ethics investigation and trailing Airbus's momentum. He was in for a year, and then out for violating the company's code of conduct. Jim McNerney was appointed the head in June 2005, by which time Connexion was actually established and seemed to have an upward pointing deployment and usage arc--I'm guessing that bought the division some time. McNerney didn't carry the sentimentality of Condit, and with Boeing on an upswing of orders, profits, and brushing away the past, must have decided it was time to take the financial hit when $320m written off would sink down among a lot of good news.

Boeing thought the business would bring in $5b in revenue per year by 2010, but it's hard to understand even that calculation. There are several thousand planes used for commercial flights in the US, but only a fraction would be appropriately large enough for Connexion to have been installed. With, say, 2,000 planes flying 300 days a year, that's 600,000 flights. With optimistic usage of 40 people on average per flight at $20 a pop, that's just $800 per flight or about $500m per year. So they would have had to have 20,000 planes or a variety of expensive services to sell airlines to reach anything like that revenue.

I was surprised as anyone to see the poor uptake numbers in this story and elsewhere: Lufthansa never saw more than 40 users on a flight, and rates were typically four percent or 10 passengers per flight. While Connexion's rate is typically reported as $27 for an entire flight, that's for the longest flights. Shorter duration flights cost less, and there were bundles and discounts available for larger companies and purchases through access aggregators, like iPass and Boingo Wireless.

On a flight that might cost $600 to $2,000 and last 6 to 14 hours, paying $27 for high-speed Internet access seems like a small fee to get a day's work back. And many regular users said exactly that. I have a feeling that outside of Lufthansa, which equipped most of its long-haul planes, there simply wasn't enough Connexion available routinely to the kind of regular traveler that would use the service. There wasn't enough for this traveler to get used to the amenity and find it out.

A comment on one blog about the shutdown wondered why Boeing didn't offer 20 or 30 minutes per flight for free to whet people's appetite. There has always been a lot skepticism about the speed the service offered, despite consistent reports that the speed was, in fact, quite good and consistent. (I occasionally received anonymous emails about gaps in coverage in certain parts of the globe, but wasn't able to confirm them.)

There's no other service in the running yet to offer global Internet access. In fact, there's no other generally available commercial Internet service in flight at all right now. AirCell will blanket North America with broadband--they will eventually have permission and partners to deploy over not just the US, but Canada, Mexico, and the Caribbean. OnAir and Aeromobile plan European launches by next year of cell service.

Gates of The Seattle Times provides a number that I haven't seen before for broadband via Inmarsat's satellite service, which is how OnAir will bring cell service into planes. Gates writes that Inmarsat would charge $6 per megabyte for data transferred. Unless OnAir has a magic deal with them, that wouldn't allow anything like the unlimited service that Connexion offered. (Connexion had enormous fixed costs, paying for transponders on satellites around the world to provide global coverage. Inmarsat is offering more of an a la carte plan used their fourth-generation satellites.)

Inmarsat doesn't yet have its third bird up covering the Pacific, which is key to making an international service viable. They claim 85-percent land-mass coverage, but that big, nearly empty ocean has a lot of people flying across it quite regularly.