MS: Neptune consumer Win2k not cancelled, just not happening
Win2k-based consumer OS slides into 'probably not' territory. Or not yet?
Microsoft has now confirmed that its Win2k development efforts have been merged into a single effort, covering both business and consumer markets, codenamed Whistler. The news, broken a few days ago by Paul Thurrott of WinInfo, effectively means that last year's triple-forked roadmap has been junked, but Microsoft's spinmeisters are still jumping through hoops to avoid using the world "cancelled." Last year's roadmaps planned for one consumer OS for 2000, Millennium, based on Win9x. In addition there was to be a Win2k-based consumer product for the 2001 timeframe, Neptune, and the next rev of Win2k for business users was codenamed Odyssey. Millennium has been substantially downscaled since then, and though it's intended to ship this year it'll create about the same waves as Win98 SE did, at the most. Odyssey and Neptune are now no more as separate projects, but although Microsoft is now conceding that the merge into Whistler has happened, it insists nothing's been cancelled. Earlier this week a company spokesman told Windows NT Magazine that it was combining and streamlining Neptune and Odyssey. This tallies with the other corporate line, that Neptune was never going to be an OS anyway, and is just a set of technologies which the company will use in Whistler. But the spokesman seems to have blown it just after by telling the paper that the combining and streamlining process will result in a product that's likely to be professional, rather than consumer. Depending on what this actually means, it might be read as throwing the Microsoft OS roadmap back into confusion just nanoseconds after it had started to look rational again. Odyssey and Neptune didn't really make sense as separate OS projects being developed in parallel, because in both cases substantial rewrites under the covers would have been needed. Bill Gates himself was saying in April last year that we'd get a Win2k kernel update this year, followed by a consumer kernel upgrade in 18-24 months. These dates would of course have been kicked back anyway because Win2k shipped rather later than Bill was hoping last April, but the important point to remember is that kernel updates were felt to be needed in the first place. The consumerised kernel you could think of as a kind of Holy Grail for Microsoft, because it's something that's obviously needed if Win2k-based code is to play in consumer markets, but it's also something that would be difficult to do, considering the footprint and compatibility issues, so actually doing it stays as an event somewhere out on the far horizon. So by merging development into Whistler and tilting Whistler towards the business end, Microsoft is setting itself a more achievable goal, while at the same time backing off from a Win2k-based consumer OS. This obviously leaves a big consumer-shaped hole in the roadmap that Millennium can't convincingly fill. The succession of minor revs to Win9x can really only be seen as stop-gaps, so that still leaves the company to decide whether to roll with the Win2k-based version at some future date (codename it Sinatra if it comes back one more time), or whether it should really do the 'exterminate Dos' rewrite of Win9x that's been rumoured (largely incorrectly) since last year. Or indeed, whether the consumer group's CE-based platforms should turn out to be the real consumer Windows. We await the next roadmap rev with interest. ® Related items: MS cancels Neptune Paul Thurrott on Windows roadmaps WinNT Mag: MS cancels plan for split roadmap