LettersNot Enough TimeAs a systems administrator for a large nonprofit organization, I have wrestled with the perpetual up-grade problem for years. While we've standardized on specific applications for word processing, spreadsheets, remote communications, etc., it's impossible to keep everyone at the same rev level.Whenever we put a computer on a desk, we have to purchase the most recent release of the application software. It doesn't matter that half of our remote users have pcANYWHERE III, we can no longer buy a compatible version of that program. Remember MS Word version 4? It was not even close to keystroke compatibility with 6.0. Keeping the MIS staff fluent in various levels of our "standard" packages consumes more time and resources than we can afford. Timothy Haight's recommendation to bring software into the public domain once the vendor stops supporting it is brilliant ("Time For 'NoughWare!" October 1, 1994, page 196). It would spur small businesses that could maintain and enhance these "outdated" packages. And, at the same time, it would allow users to upgrade when we decide to-not when the software vendors say it's time. Joseph Pescatello
Upgrade v3.$If software makers had 'NoughWare ("Time For 'NoughWare!" October 1, 1994, page 196) they would lose all that instant revenue they get with an upgrade! I wonder how much Microsoft made on the DOS 6.0 to 6.2 upgrade? Don't you think that the same thing is happening in hardware? How about 'NoughChips now?I am a department manager and have involve ment in organizational information. This is a small rural hospital and many of us wear many hats. That's why I find your publication so useful. It helps keep us informed. It helps us to make decisions and plan for the future. Thanks again and keep up the good work! Mark Perry
On The Same WavelengthPerhaps it is the type of misinformation distributed by so-called network experts such as Bill Frezza, specifically in "Cyberurban Redevelopment," (October 1, 1994, page 39) that leaves the uninitiated with the belief that all information systems professionals watch too much Star Trek (or others in the genre of science fiction).One of the most basic concepts in television signal engineering is the interference contour of a station --the fact that a broadcast occupies not only a particular frequency (or, more down to earth, a channel number), but also several adjacent ones (most interfere with approximately eight in the UHF band). This explains, with the exceptions of Channel 4 and 5 stations in the VHF band (between which there is a frequency gap), why there are never any stations located immediately next to each other in the same band. It also, in broad terms, explains why poorly designed, first-generation Pentium machines are able to block out the Rolling Stones on your radio. But, even with the widespread implementation of cable TV in most areas, it's refreshing to know that I can still get Channel 20 in Washington. So, until you can convince IS professionals that a 500-Kbps wireless LAN is more important than reruns of Star Trek, don't hold your breath for the space in the UHF band. Timothy D. Jasionowski
Bill Frezza replies : Thanks for reminding everyone how notoriously inefficient the NTSC broadcast television standard is-that 50-year-old technical compro mise that has frozen us in time with the world's worst TV picture quality. When this standard was adopted, low-cost RF receivers had filter characteristics that pale in comparison to what we can achieve today. Yet, here we are, stuck with adjacent channel interference rules that no longer make any technical or economic sense. All the more reason to get the UHF TV industry to clean house. And as far as deciding what is more important-old Star Trek reruns or wireless LANs, why don't we let the market decide, especially since it easily can have both. In fact, I suspect your employer, MCI, which is currently forking over close to 40 percent of its revenue to the local phone monopolies, might find even better uses for that spectrum. Timothy Jasionowski responds: Bill, didn't say I disagreed with you. I basically said that the interference contour should have been noted while you used your channel-changing reference -- kind of misleading. And by the way, I'm IS manager in MCI Access Management and reminded each day about how much we fork to the Baby Bells. CORRECTIONSWe misprinted the name of Znyx Corp. in our November 1, 1994, PCI Adapters review ("Ethernet Cards Take the PCI Bus to Better Server Performance," page 98) and the 32-Bit NICs Buyer's Guide chart (September 1, 1994, page 193). We apologize for any confusion these errors may have caused.In our October comparison of remote access servers ("Remote Access Without the Hassle," October 1, 1994, page 86), we mistakenly stated that Xyplex's MAXServer 1620 did not have a GUI management system. Xyplex does offer GUI management by way of ControlPoint, a SNMP Hub network management system. Xyplex states that ControlPoint provides complete configuration and management of the entire MAXServer product line. In our November 15, 1994, State of the Technology issue, we incorrectly identified the photographer of the picture of Gary Stimac, senior vice president, Compaq, that appears on page 98. The photo was tak en by Pam Francis. We apologize for the error. |











