MS keeps mum on NT

Microsoft: the news is there is no news. The Register: Eh?

Microsoft has, unusually, begun to restrict the release of information regarding its Windows strategy. Normally Microsoft press information is piled so high that it creates an obstacle course, but strangely, 18 hours after the keynote speeches began at the WinHEC in Los Angeles, there are no promised transcripts of the speeches, and nor was there a Webcast as is normal for such events. For some strange reason, The Register's invitation to attend must have been delayed in the mail. The conclusion must therefore be that Microsoft is seriously embarrassed at its Windows plans, and wants no clear roadmap to emerge. On the NT front, there are supposed to be consumer, workstation, server and enterprise versions. There are also reports of a forthcoming embedded version, based on NT4.0, which needs between 8MB and 12MB of RAM. Some bed for this embedded version: it looks as though Microsoft will be producing embedded NT for semi-mobile mobile phones, or luggable networked toasters that require the man of the house to lift it off the shelf. Just to make life interesting, this embedded NT uses extensions that are not part of regular NT, because the design of NT is deficient. Isn't it strange that Dave Cutler, the original architect of NT, is not featured at the WinHEC to receive the plaudits of the assembled multitude? Reports suggest that the embedded version will be released to manufacturing at the end of the year, a new tactic to mean it will appear three months later than that, at the earliest. It's supposed to be being tested at 500 sites, but we are inclined to view such claims with suspicion without supporting evidence, in view of the deviousness of claims that we have all seen during the Microsoft trial. Delays to software are normal and expected. Microsoft's problem, and a very serious one now, is that it has raised consumer and business expectations prematurely, and caused planning havoc for PC makers. It is quite extraordinary that Dell and HP are reportedly planning to preinstall the Windows 2000 beta 3 (aka NT5) later this year, as though this were the real thing. It reminds us of the days when Microsoft used to issue, to the media faithful, reviewers' guides for beta software. The latest reports about NT5 Server is that the clustering does not work, so perhaps those who need it should continue with tried and tested Unix, even if it isn't sexy nowadays. After all, if Cobol code can last for 30 years and be resilient enough to cause the Y2K problem, why shouldn't Unix be good for another 30 years too? We can only conclude that the WinHEC has come at a bad time for Microsoft, and that the strategy is not to divulge a strategy. ®

Sponsored: 10 ways wire data helps conquer IT complexity