They did not invent mobile computing, they invented a shop

In March of last year I wrote that mobile software would stay worth roughly nothing until somebody controlled both the handset and the place you buy software for it, and that whoever managed it would look like they had invented mobile computing when what they had actually invented was a shop.

That happened this summer, and a second one arrived last month, and I want to be clear that the shop is the whole story.

Nothing technical changed. Phones could run third-party software before. What changed is that a developer can now write something, submit it once, and have it appear in front of every owner of a device, with the payment handled, without negotiating with a single network operator.

Compare that to what it replaced. Approach each carrier separately, in each country, submit to each approval process, accept whatever placement they offer, and hand over sixty or seventy percent of the price. Then port to hundreds of handsets that each interpret the runtime slightly differently.

The new arrangement takes thirty percent and one submission. The people who spent a decade calling that split greedy have either not done the arithmetic or were not in the previous business.

What I got wrong is how fast it would matter. I expected a slow build. Instead there are thousands of applications within months, because the barrier was never that developers lacked interest. It was that reaching a customer required a commercial department, and now it requires an upload.

The part worth watching from here is that a shop with a single door has an owner, and the owner decides what gets through it. Approval is discretionary, the criteria are not fully published, and there is no appeal to anyone outside the company. Nobody minds this much yet, because the alternative was so much worse and because the rejections so far have been small.

They will not stay small. A platform that becomes the primary way software reaches people will eventually reject something that a lot of people wanted, for reasons that look commercial rather than technical, and the argument about who should hold that power starts on that day.

Carriers held exactly the same power for a decade and used it worse. That is a defence of the new arrangement, not of the concentration.