Beyond OS development: what's next for Microsoft

'Twas the night before Teched, and I had just installed Release Candidate 1 of Windows XP Professional. At the RC stage, it is really all over bar the shouting. Changes can be made, but must be cataclysmically important to get through the testing and QA process in time for final code shipment.

I sat back and considered whether it was worth it: whether XP introduced new capabilities that would justify it in the marketplace, and whether network managers should be coughing up their hard earned departmental pennies to buy the upgrade.

The NT 3.1, 3.5, 3.51 provided new capabilities and reliability. NT4 was a bit wobbly at first, but settled down after a few service packs. Windows 2000 was a big leap forward, and benefited from two service packs.

XP is easier to use, and the new UI is pretty and effective. It is also "a better Windows 2000". Microsoft has achieved almost Macintosh-levels of installation ease of use with XP - just shove in the CD, run the set up program and let it work. It kills off the Win9x legacy base, but surely no one with an ounce of business sense is still using that platform today.

The difficulty will be with the server. Mistakes were made in the implementation of Active Directory in Windows 2000. The promised inclusion of significant .Net functionality might well be the kick this platform needs to propel it forward. It's not that it doesn't work - it does. But should we really be working with tools like Office?

Isn't it all rather 1980s? With each release of the operating system platform, Microsoft has solved a significant problem area. It's difficult to see where the company can go from here.

More hardware support would be nice. A funkier desktop design perhaps? Maybe. One thing is certain, XP draws a line in the sand heralding the end of a long period of operating system development.

So where will the next leap come from? The time is ripe for a push on the management front, both of knowledge and infrastructure. Microsoft's portal technology, codenamed Tahoe, offers a lot of capability but is crying out for full Active Directory and .Net integration.

The same goes for Microsoft Operations Manager. The word is that Kodiak, the codename for Exchange Server 2003, will move storage from Exchange JET to SQL Server. Maybe the elusive common object store will finally come to fruition, with management and knowledge tools that allow you to locate, work with and manage all information in a coherent manner.

This is the next big nut that needs to be cracked, and Microsoft has all the core technologies in Release 1 products. That's why I look forward to Windows XP +1 - it promises to be a vintage release.

Jon Honeyball is a software developer and computer consultant.