It’s interesting to consider the history of the PC and learn some lessons. The landscape of computing that we have now resulted from a sandwiching effect of cheaper and cheaper scientific workstations and more and more powerful consumer hobby kits (some say toys). By the mid 1980s, this trend was recognized and efforts were begun from both sides to capture the computing market. Today we know that the hobby kit lineage won, and as a result, most workstation companies eventually folded in the 2000s, though they survived for a while by clinging to the enterprise. (Incidently, IBM and HP did not, because they were large and diversified enough to do something about it.)

But this is not the whole story.

Look at how hobby kits finally became respectable: not with the ad-hoc hacked-up first iteration legacy which resulted in a 20-year detour through DOS/Windows 3.0, Apple II/Macintosh. People often say that the PC revolution began in 1975, but all the systems commonly used today do not go back that far, they trace back to de novo R&D from the late 1980s / early 1990s that attempted to emulate workstations rather than make toys. Besides NeXTstep, which became Mac OS X, there was NT OS/2, which later became Windows 2000. Even GNU/Linux, which is the most UNIX like of all, came out with its first distribution in 1993. This is remarkable. It means that toys could take over the world, but cheap hardware alone was not enough, blindly ported software wasn’t enough, hacked together software wasn’t enough, it had to be well-engineered software made specifically for the device and its new usage scenarios — something that wouldn’t have been known until people played long enough with the toys.

So today, we have mobile devices that are still consumer toys for the most part. But like an old movie, the gap between mobile devices and PC’s is closing; the case for PC’s and PC companies is increasingly enterprise customers; and attempts to shove a downgraded PC directly into a mobile form factor like Blackberry/Windows PDA’s and netbooks are not working as well as making toys more powerful. Some things don’t change. New usages like electronic payment, geolocation, and image search are clearly emerging from the toys, and there will be a second iteration of development that make mobile devices seriously useful. It won’t be through making them do all the things that can be done now on the PC though — that would be impossible — it will be through making them do things that make things now done on the PC obsolete, because there will be better ways. The next ten years should be interesting.