Microsoft junks NT-based consumer OS plan

Convergence after Win98 has been 'postponed,' apparently...

Microsoft will produce at least one more rev of Windows 9x before it finally converges its operating systems, it has emerged this week. This new plan flatly contradicts what the company has been saying for over a year - far from rolling its desktop products into one basic OS at Windows 2000 stage, Microsoft may now be on the point of abandoning its convergence strategy. Microsoft has actually been talking about rolling it all into one product since the early stages of NT. For most of that time the company hasn't got any closer to a single core product that covers both business and consumer sectors, but more recently it has made it clear that it sees Windows 98 as being the last of the line. The next big product, Windows 2000 (aka NT 5.0) was intended to spearhead the push to move NT technology into the corporate market, supplanting 9x, and also to form the basis for a new consumer-oriented operating system. But now it would appear that all of this has been at least postponed. Microsoft is planning a Windows 98 service pack for the middle of the year, and then after that a more substantial 9x update aimed at the consumer market. Considering the calendar and Microsoft's historical problems in getting new revs of its operating systems to market, Windows 99 isn't an option as a name, and it appears the company is currently referring to the product internally as Windows NT Consumer. Aha, you may note. Considering that NT is going to be called Windows 2000 next, NT Consumer obviously isn't going to be the name either. As you read, focus groups are no doubt busy on this little conundrum. There are several probable reasons why Microsoft has decided to 'delay' convergence. Windows 2000 itself is a moving target, and it will be virtually impossible for MS developers to get going on a consumer OS based on the NT kernel until Win2k has shipped. It's also a big, unwieldy monster which - in order to sell to its business target market - is going to have to be secure and stable. It probably isn't where you'd want to start if you wanted to develop a consumer OS that was good at games. Essentially, Microsoft's developers will find life a lot easier if they stick with the 9x kernel rather than continuing to try to reverse stuff into NT, compromising that product as they do so. But we can spot what's happening here, can't we? That is precisely the reason Microsoft's vaunted convergence plans haven't got anywhere since the early 90s. The idea might work in theory (actually it doesn't but never mind that), but the reality is that it's impossible to do in practice. Note also that the Windows NT Consumer tag, even if it's unlikely to ship under that banner, bears a certain relationship to other Microsoft 'Windows' projects. Rather than converging Windows operating systems, Microsoft has found itself adding new ones as it goes along, and in the case of Windows CE, even splitting these into more variants. The product name in general expresses a belief that Windows is whatever we call it, and maybe Windows is just about a common look and feel (although that goes out the, er, window when we're talking appliance or in-vehicle implementations). Convergence is something Microsoft wants to do, sometime - but it'll probably never really do it. ®