OS UPDATE - Part 3


by Wayne M. Krakau - Chicago Computer Guide, October 1998

Here’s the third in my series on the latest operating systems issues. This article completes my coverage of the upcoming release of Windows NT Version 5.
To continue with my ramblings on NT 5, another hurdle that it must overcome is the potential conflict with its customer’s Year-2000 compatibility efforts. With all of the expected delays, NT 5 is expected (as opposed to announced) to be released in mid-to-late 1999, just as most companies are going on full Y2K alert. Programming and support personnel are scheduled to be working overtime, trying to keep up with any last-minute Y2K-related errors.
Remember, Y2K errors can pop up at any time a calculation looks ahead beyond January 1, 2000, not just after that actual date. As we approach that magic date, applications that calculate receivables, payables, loans, leases, individual or project schedules, court dates, or just about any other date-related fact are eligible for failure, or worse yet, subtly inaccurate data. With all of this to worry about, it’s no wonder that companies are less than anxious to make a major upgrade to their network and desktop operating systems right in the middle of this mess.
Now, word is out that Microsoft’s own closest corporate partners are rebelling at the idea of doing simultaneous Y2K and NT 5 implementations. These are the major corporations that would normally have agreed to unconditionally implement NT 5 immediately upon release after having been top-level beta testers, all in return for major price discounts and the promise of greatly enhanced support services from Microsoft.
Microsoft can’t guarantee a specific release date, but expects these companies to agree to implement NT 5 en masse on whatever day Microsoft chooses. This issue has yet to be sorted out, though I expect that many of these corporate partners may jump ship, resulting in major embarrassment for Microsoft’s marketing department, and a major headache (in form of less feedback from the trenches) for Microsoft’s techies.
My biggest fear here, is that Microsoft will be tempted to do with Windows NT 5 what it did with Windows 98. That is, it arbitrarily picked a date and decided that for marketing purposes (or as they say in the Far East, to save face) they would simply release their product, ready or not. With Windows 98, one week there were published reports of a release date six months in the future, with lots of bug-fixing work left, and the next week the software was released! Windows 98's reliability has suffered greatly from its rush to shipment, and I’m sure that NT, being an inherently more sophisticated product, would have even more problems as the result of such a premature release.
Another issue is, as NBC so eloquently puts it, is the "It’s new to you" syndrome. A particularly difficult to implement technology such as a network directory services system, takes much experience to get right, and if you haven’t done one before, as in Microsoft’s case, then it’s new to you. That means you’re a beginner, likely to make all of the mistakes that anyone would reasonably expect from a beginner.
When Novell was a beginner in directory services, they came out with NetWare 4.0 (along with its descendants, 4.01 and 4.02) which was elegant in conception but a real disaster in terms of actual implementation. They aggravated things by releasing it before it was really ready, just as I am afraid that Microsoft will do with NT 5. NetWare 4.0 was so bad, in fact, that I declined to offer it to my clients, and actually gained a few new clients when I was called upon to uninstall it and put back a previous version of NetWare for some corporations.
Novell didn’t get things right until they released version 4.1. That product followed up on the promises inherent in the design of 4.0. As is usual for Novell marketing campaigns, no serious effort was made to explain to the public that 4.0 was a mess and that 4.1 was the "really working" version and was radically improved as compared to 4.0. That’s why, to this day, I still run into applications developers and sometimes even civilians (non-computer folks) who are afraid to move from NetWare 3.x to 4.x for compatibility reasons. They fear that their application programs won’t work with anything higher then NetWare 3.x.
I have no reason to believe that Microsoft, which already has a shaky reputation as far the reliability of major software releases is concerned, should magically be able to create a stable directory services system on their first try. Of course, there is always the hope that they have learned from Novell’s mistakes, but so far, they haven’t even learned anything from their own, so don’t hold your breath.
Whatever happens with NT 5, I am resigned to the fact that, due to marketing-driven customer demand, I will probably have to sell and support it. As long as I don’t get blamed for its problems, my company will survive and probably thrive on the increased number of billable hours.
Next month I’ll cover NetWare 5. Notice that I didn’t say "the upcoming release of" NetWare 5? That’s because I just received E-mail that notified me that NetWare 5 has been released more than one week early! Wow! How’s that for innovation? What will those people at Novell think of next? Maybe they’ll even start marketing their products properly - or is that too much to hope for?

�1998, Wayne M. Krakau