October 1999

Input/Feedback

The following letter refers to Tom Keating’s Cc: Column in the August issue of CTI�, “Windows 2000 — The Next Great CTI Platform?”

I am new to the CTI industry, though I’ve been working extensively with computers for more then 20 years, as a VAR and a consultant. In every account of Windows 2000 that I have seen, the point was clearly made that the one of the key issues with the OS is its reliability, that Linux or Unix are far more stable, and that few IT pros are willing to put mission critical software on an NT or Windows 2000 platform. Yet here I read an article that could be in a brochure handed out at a Microsoft booth at a trade expo.

It’s not until the second to last paragraph that he offhandedly mentions that “today’s flavor of Windows NT (version 4.0) lacks the four nines of reliability (99.99 percent) that many businesses require.” First of all, the number really bandied about is five nines (99.999 percent) uptime for mission-critical systems. Again, I’m from outside the fold here, but there’s just no way I would ever recommend a Windows OS as the basis for something as critical as phone systems, at least not without some serious caveats.

Robert Fately

Tom Keating replies:

My column was not a full-fledged test-drive of Windows 2000. The column’s purpose was to report on the latest features and new functionality that Microsoft is putting in their next major operating system release, which will have an impact on several industries, including the CTI industry. It is my opinion that Windows 2000 is an important milestone, and will have an impact on low to mid-range server environments. It was my sole intention to report on the advantages of Windows 2000 and Microsoft’s attempt to crack the mission-critical market.

I would have to strongly disagree that my column did not address some serious limitations in Windows 2000. Not only did I point out that Windows 2000 does not have four (or even five) nines of reliability, I also mentioned that I was able to crash Windows 2000 in just 15 minutes. Being too critical of a beta product before the bugs are worked out and the software is released is not good journalism.

One final comment regarding whether Microsoft helped me write the article. The answer is an unequivocal “no.”