A story on digg.com reports that a computer cluster running OpenVMS as its operating system has been running 10.5 years without crashing. The same day we see speculation that Apple will release an iPod by the end of the year with 16 GB of flash memory.
It’s an interesting juxtaposition: Remarkable performance by old technology, performance in terms of uptime that is irrelevant for the vast bulk of present computer-based devices (in terms of quantity), no matter how exotic they get, and a reminder of the transience of the personal computer.
Sadly, the backbone of most corporations’ IT, the desktop PC, is stuck in a technology dead end and will benefit neither from the innovation that powers the price-performance of consumer electronic devices, nor the technology that provided a core characteristic of the highly developed mainframe and minicomputer world, namely reliability. Consider, for example, the mobile phone. Steady innovation has reduced its size and price and dramatically extended its functionality: MP3 ring-tones; built-in video cameras; and Internet access, for example. As innovation shifts to smaller and smaller devices, such as wireless sensors associated with an emerging Internet of things, the PC becomes even more of a backwater technology.
Of course, the hardware innovations, and to some extent the software innovations, such as codecs, trickle down to the PC market, but mostly the PC, serving a totally different need from that of phones and iPods, is unaffected.
While the hardware becomes increasingly reliable, and software becomes translated into firmware, we put up on a daily basis with PC-based software that breaks.
The heart of this problem is the operating system. It is the nature of monopolies to impose what economists call “externalities” on the world, i.e., costs that aren’t accounted for in the production and use of the product. In the case of Windows the chief externality is the time we all waste waiting for Windows to reboot, or for some third-party applications to unfreeze. As the story that leads this column indicates, it is not as though the technology for operating system reliability did not exist: It was just ignored or omitted.
With Apple almost completely out of the picture on the desktop, with a measly three per cent of the market in spite of being so cool, there is little market incentive to change the nature of the PC. Nor, in fact, is there any particular technology incentive to do so. It makes so much more sense in a free-bandwidth world to strip the operating system out of the PC so that it becomes a comparatively simple device running applications hosted on ultra-reliable server farms (the way we experience Google today). Until this perfect day dawns we can only read stories about what might have been, and wince. The PC could have been so much less painful.
Tim Warner is an independent strategy consultant focusing on the intersection of IT developments and business strategy. firstname.lastname@example.org