An Unfair Comparison?

I don't know much about database software or database servers, so to broaden my knowledge, I read articles such as "Justifying NT" (April). Unfortunately, the narrow scope of the article and the explanation of test results described on pages 150151 cause me to wonder if the information is reliable. The testing is performed using OS/2 1.3 because "Microsoft recommends OS/2 1.3, rather than OS/2 2.1, as the best operating system on which to run it." Is it not the responsibility of BYTE to test the manufacturer's claim? In addition, results for Query 2 are not shown because "NT would not process this query." Perhaps it would be of interest to see the numbers for those systems that did manage to perform the task. One wonders how NT can be viewed so favorably when it cannot process a list of 100 orders.

Richard Hodges

Los Angeles, CA

For 16-bit OS/2 appl ications, version 1.3 is a faster, leaner platform. However, the point of the article was to assess how one RISC-capable operating system, NT, can and cannot fit into an enterprise setup. The conclusions were decidedly mixed: NT can be a good applications server, but don't plan to toss out your file server and put in NTAS. Funny, you don't claim it was unfair to slam NT for its lousy performance on the file-server tests on a token-ring LAN where it lacked 32-bit drivers, or how we showed that OS/2 was superior there.--Jon Udell