Justice throws Microsoft OS/2-shaped lifebelt
Did Microsoft cynically destroy OS/2 by, er, paying IBM to employ all those useless execs?
Has the Justice Department thrown Microsoft some handy red meat in the shape of John Soyring? The IBMer, now director of Network Computing Software Services and formerly in charge of OS/2 development, took the stand yesterday, but the appearance of the O word could be exceedingly helpful for Microsoft. Soyring pointed out that restrictions in Microsoft developer contracts limited the ability of developers to produce applications for non-Windows platforms, and somewhat exquisitely pointed out that giving customers the ability to choose their own browser and to modify the user interface hadn't harmed OS/2. Well, yes… But the catastrophe that was OS/2 has so many IBM fingerprints all over it that it will be virtually impossible for the prosecution to establish that it was all Microsoft's fault. And so far, Soyring's testimony hasn't gone anywhere near the most obvious Microsoft sins from that saga. Microsoft can reasonably argue that it's none of its business to bankroll rival products by allowing developers to use its tools and code to develop apps for them. Soyring argues that OS/2 was impeded because it didn't have critical mass of apps, whereas Windows succeeded because it did have this. Microsoft meanwhile on the court steps argues that OS/2 failed because it was too resource-hungry. But the two rivals did start from approximately the same point - there weren't many Windows apps around when OS/2 2.0 (the big one, as was intended) first shipped, and OEMs did not as a matter of course ship Windows with their machines. Microsoft did an excellent sales job on the OEMs and created the momentum for application development, while IBM did not. Microsoft's methods may or may not have been robust, IBM's abilities however were demonstrably feeble. The parts of the Microsoft-IBM relationship which might be worth examining, but which don't appear to be being looked at, concern the nature of their agreement prior to the breach, and Microsoft's 'ambush' of OS/2 2.0 with Windows 3.1. Microsoft was as far as their agreement was concerned supposed to be carrying on running with Windows as an entry level system. Users would then migrate to OS/2 as the system matured, and as it became more realistic from a hardware real estate point of view. Microsoft was also supposed to be developing Portable (i.e. non-Intel) OS/2, but wound up developing a rival product, NT, instead. Basically, it stabbed IBM in the back. Then with the ambush it stabbed it in the front. When IBM rolled out OS/2 2.0 with Windows 3.0 support, Microsoft moved the goalposts to Windows 3.1, denying OS/2 2.0 critical application support early in its career. Microsoft did of course have to persuade developers to switch over to the new version fast (and actually, get developers on board, because it didn't have many then), but it succeeded. IBM, on the contrary, conspicuously failed to win developer mindshare despite (to some extent because of) several years of different developer programmes, shifting market positioning and shifting strategies. Career-wise, it may be helpful for IBM execs from the period to blame Microsoft, but the long list of IBM screw-ups points to an entirely different explanation for the failure of OS/2. ® Complete Register trial coverage