News of future Windows rocks the Net

A fascinating report by ZDNet's Mary Jo Foley sent the Web into a tizzy overthe Fourth of July weekend, with its story of two new Consumer Windowsproducts that Microsoft hopes to release in 2000 and 2001. I normallywouldn't have commented on this article, but I've gotten so many requestsfor clarification that I figured I should chime in with my own explanationof the news.

The story goes like this: In 2000, Microsoft will release its Millenniumproject, a Consumer Windows that is based on Windows 98, not Windows 2000.Millennium, which has yet to be officially named, will be the cornerstone ofthe "EasyPC" initiative, an attempt at making a new kind of PC that doesn'trequire the user to open the case to add memory, new hard drives, or anyother kind of hardware. In an EasyPC machine, expansion is handled viaDeviceBay cartridges, which will act like video game cartridges, and byexternal devices such as FireWire (IEEE 1394) and USB. Other than that,little is known about Millennium: It is expected to include a vastlysimplified user interface and be geared toward typical home use scenarios:Game playing, Web browsing, Internet email, and the like. None of theMillennium information in this article was news to me. In fact, I'vereceived numerous emails from beta testers who were recently contacted byMicrosoft to join the Millennium beta, so a late 2000 release for thisproduct is indeed possible.

The ZDNet article, however, also discusses a post-Millennium version ofConsumer Windows that is code-named "Neptune." Like Millennium, I've heardabout Neptune; John Dvorak supposedly released information about thisproduct earlier this year (he didn't, actually, but that's another story),showing off its HTML user interface and communications-based design.According to Foley, Neptune will follow Millennium in 2001 as the newConsumer Windows, but will be based on the NT kernel used in Windows 2000,not the 98 kernel. And Neptune will be the first Windows OS to feature the"WinTone" promised by Bill Gates at Fall Comdex 1998. With WinTone, Foleywrites, Windows PCs will be self-healing and self-repairing. Of course,Windows 2000 already features this, so I'm not sure what WinTone adds to theparty. I suspect it has more to do with 24/7 Internet connectivity thanthose other features.

But there's more: Neptune is designed to be the ultimate client on aUniversal Plug and Play (UPnP) network, where devices auto-configurethemselves for the network and broadcast information to other clients thataccurately identifies themselves. Neptune will also form the basis ofpost-Windows 2000 business OSes from Microsoft, so Microsoft is encouragingthe Consumer Windows and Business Windows to get together and decide on acommon user interface and other components. Perhaps most intriguingly,Neptune may be used as the basis of non-PC devices that will likely be basedon EasyPC: Microsoft is considering media devices (for browsing the Web) andgaming consoles that are based on Neptune.

The biggest challenge in moving Consumer Windows to the NT/2000 core, ofcourse, is compatibility. Microsoft had originally wanted Windows 98 to bethe end of the line for the 9x line, but problems getting Windows 2000scaled down to their consumer customer's needs forced a rethinking of thatstrategy. By the time Neptune actually ships, the industry should have movedon to a new type of device (EasyPC and its relations), making the need forlegacy (ISA, DOS, etc.) support superfluous. The key goal for Microsoft,then, will be application software compatibility: Currently there are fewentertainment titles that run successfully on Windows 2000, compared toWindows 98. Over the next few years, refinements to DirectX and the Win32API should eliminate these problems. So, by the time Neptune actually doesship (and, no, it won't be in 2001), compatibility issues should have beensolved, if only by the passage of time.

And that's really the only problem I have with the scenario reported in theZDNet article: The timeline. Anyone who believes that Microsoft will ship aconsumer OS in 2000, only a year after Windows 98 SE, and then ship anotherone in 2001, only a year after Millennium, is being delusional. To be fairto ZDNet, they're simply reporting on internal Microsoft documents: Thedelusion here, of course, is in Redmond, which still believes it can shipproducts on schedule, despite a decade of evidence proving otherwise.Windows 2000, for example, has been in beta for almost two years. There isabsolutely no way they're going to ship two consumer OSes, two years in arow. Fuggetaboutit.

If you were wondering whether I was going to refute the ZDNet article,wonder no more: To my knowledge, there's nothing wrong with it (other thanan inane and inaccurate quote from "betanews"). However, I should say thatmost of the Neptune information was news to me so I can't vouch for itsaccuracy. I suspect that the next few years are going to be quiteinteresting for Windows users and I would like to offer up a cautionary bitof advice for anyone thinking of moving to Windows 2000. It goes somethinglike this:

When Microsoft decided on the name change from Windows NT 5.0 to Windows2000, the thinking was that the 9x and NT products lines were being merged,so a single product called "Windows" made sense. However, this decision wasa poor one: NT wasn't ready to take over the Consumer Windows product line,as it turned out, and customers used to products such as Windows 95 andWindows 98 naturally assumed that Windows 2000 was the next obvious upgrade.It isn't. Folks, Windows 2000 is a business OS, designed for corporateservers, desktops, and mobile users. Despite some concessions toconsumer-oriented technologies such as DirectX and 3D display adaptersupport, Windows 2000 is not for home users. It's not for game players. Andit's probably not suitable for most of the people pining away for itsrelease. Windows 2000 isn't as compatible with existing hardware or softwareas Windows 98 is. And now that future Windows such as Millennium are on thedrawing board, hopefully a large number of the people previously looking toWindows 2000 will now have a new goal for their next upgrade.

That's not to say that Windows 2000 isn't superb: It is. But its biggeststrength is its NT core, which isn't necessarily appropriate for mosthome/consumer users. Yes, there will always be the so-called power usersthat want to run Windows 2000 Professional (just as they're now runningWindows NT 4.0), but for most people, Windows 2000 isn't the solution. Butnow you know what the next few years look like and you can plan accordingly.Just add a year or two to the previously stated timelines when you startthinking ahead: None of this stuff is going to happen on the schedules we'veseen here of course