I have Microsoft's Jim Allchin on the phone, and in my head I'm counting up the number of Windows versions I'm running at home. There's the old 3.1 system running WordPerfect, which never crashes and is a favorite of the kids for typing. There is the 98 machine, which crashes if the cat walks by, and the ME system, which is slightly better than 98 but not much. And, finally, there is the Windows 2000 system, which runs solidly but seems to decline any game the kids want to throw on it.
Now Jim is saying that all this code base will come together in Windows XP. "I am really pumped over it," Allchin said, championing the system's crash-proofing, multimedia extensions and voice capabilities.
I give Allchin credit. He pushed through the Windows 2000 project when doubters, including myself, said the code base was too big, too unwieldy and too bug-ridden to ever make it out the door. He's also one of the increasingly few longtimers at Microsoft. He has dimensions outside of code bases and bug fixes, including an interest in guitars and blues music that can be refreshing in a PowerPoint-driven culture. He has also never launched into a diatribe on how Banyan VINES "coulda been a contenda," despite his history as Banyan's chief technology officer.
But I don't think crash-proofing, multimedia stuff and voice is sufficient to remake the desktop world. Championing a new version as less crash- prone than the previous version always seems to me a perilous course. A bit like saying the new Yugo model rusts out much less frequently than the previous year's. Linux doesn't crash, Apple didn't used to crash (I don't know about the latest rev), and my toaster doesn't crash unless I toast my bread during a California rolling blackout. Let's just leave it at this: Any new operating system should be able to run for weeks without failing.
Multimedia matters far less now that Napster is napping and the excitement of pressing the entire collection of Lothar and the Handpeople can get you a couple of weeks at the county farm. With the telco companies falling over themselves to cut rates to zero while throwing in a free phone, it is tough to see a wholesale shift to phones over IP services even if they do perform decently.
So what does matter? I'd say the first stop is security. Windows is ubiquitous on the desktop and the hackers' favorite target. A 10-bullet point slide on how XP is going to resolve the issues of rogue attachments, unhindered access to your Outlook address book and backdoor defenses would go a long ways to having folks think an upgrade is a worthy investment. The other issue that I think would be well worth the development dollars would be power-saving features that a user can actually understand and adjust via a graphical interface. I don't want seven power options from which to choose. I just want the system to be secure and powered down when I am done for the day.