A decade or so ago, my partner's cell phone provider was bought by AT&T and the old network was to be disconnected. AT&T's network was incompatible with their existing phone so they were required to get a new one.

The only smartphone they could get for free was a Nokia device running Windows Phone 8, so they picked that.

Their level of technical sophistication was not very high and this was to be their first pocket computer.

It had a fraction of the CPU grunt of my Galaxy S5 so I expected it to be slow and for them to hate it. I also expected to be asked to solve problems with it and help them along with some aspect of it or another.

But there was none of that. It just worked. They never had any questions. Like many people with a pocket computer, they came to use it all the time for things.

I poked at it myself a few times and found the user interface to be very different from Android and IOS, but it flowed well and was always instantly responsive. It was a neat little machine that seemed to perform extraordinarily well.

And despite finding a way to get this kind of positivity from me, a former OS/2 zealot and long-time user of free operating systems, they still managed to completely fuck up the entire operation. It remains the only example of a Windows Phone device that I'm aware of ever having seen someone use in the wild.

I think what really held them back was that Wi-Fi was only starting to roll out, and outside a hotspot area, the universe of things you might do with one was necessarily quite self-contained. It limited what “killer apps” could be developed, as anything designed for the platform probably needs to be fully offline most of the time.

Too bad the tooling around it was so bad. I should do a writeup of why, it is an interesting case study in how poor extendability of tooling can hurt an entire company.