The joy of teaching. Remember that Pentium-75 that was limping along under Windows NT’s heavy yoke? She didn’t complain to me about it (probably because she knew I did a lot to try to make it usable), but she did complain to some other people. One of the other IT guys did some lobbying. And when I said that a Mac would be an improvement over that thing, it got some people’s attention. (I’m not exactly known as a Mac zealot at work. Some call me exactly the opposite.) So she got another machine.
Well, as it turns out, she’s taking a class titled Management of Information Systems. She called me up yesterday to ask me a few questions relating to the class. Sure, says I. She asked about a mainframe’s place in a Webcentric world, which really made me think. I’m not of the mainframe generation, and I left the computer science program at the University of Missouri because the only thing they were interested in cranking out at the time were IBM System 370 administrators who knew VM/CMS and JCL. Gag me. I’d rather use and fix Macs. But retrofitting certainly makes more sense than outright replacement in many cases.
Then she asked about NCs. I laughed, because my now-defunct Linux book was to have a chapter about NCs in it (and how to roll your own). “Why didn’t they catch on?” she asked. Two reasons, I said. Poor marketing, for one. Larry Ellison assumes that everyone hates Microsoft as much as he does, so he releases this overpriced box and says little about it other than “Not Microsoft.” The second reason, of course, is versatility. People like the versatility of their PCs. NCs have none.
Then she asked a sharp question. “Isn’t this the same thing we do with Reflection?” (Reflection is a very high-priced VTxx terminal emulator from WRQ, Inc. that we use to connect to a cluster of VMS boxes.) Ah, she gets it! Yes, only NCs pull Windows and Windows applications (or another GUI and GUI applications) instead of text-based programs.
“This stuff doesn’t even seem real,” she said at one point. “And here you are, talking right off the top of your head about it.” But at the end of the conversation, she seemed to get it.
And that’s what’s cool about writing books or maintaining a Web site. Lighting up the darkness. Making the unfamiliar make sense. Or at least a little more sense.
My phone was ringing off the hook today. I made a comment about my popularity rising. One of my office-mates suggested I run for president. Well, I said, I couldn’t do any worse of a job of carrying Florida… But I won’t be of legal age until 2012.
I guess that’s the last thing I have to look forward to. At 16 you can drive. At 17 you can get into R-rated movies. At 18 you can vote. At 21 you can drink. At 25, your insurance rates go down. And at 35, you can run for president.
I remembered seeing an article a while back concerning this person’s issue:
It is an optimization guide for the K6-2+ (also K6-III+, but not
explicitly stated) processors, and it includes a board compatibility guide.
According to the guide, the FIC VA-503+ will only support the new
processors with a beta bios (doesn’t mention specific versions) and revision 1.2 of the board.
I’m CC’ing a copy of this to Curtis.
Dustin D. Cook, A+
I didn’t get this message, except when Curtis replied to me. I’ll have to investigate.
Thanks much for the tip.
Doesn’t anybody else feel impelled to mention that a 50MHz (for in this case >12.5%) speed gain is completely valueless? As you surely know, Dave, in ordinary use people don’t usually notice any speed change that’s much finer than 2x. I’d take a 12%-faster CPU and pop it into my system if 1) somebody gave it to me for nothing, and 2) it was a no-brainer plop-in install. Otherwise, there are better uses for $60. Ya think?
Peter A. Moore
Precision IT, a division of Precision Design Systems
in reference to:
My understanding was he wanted to get that CPU because he was using his 400 MHz CPU in another system.
Yes, you are entirely right, a 50 MHz gain generally isn’t worth it. You could make an argument for when it’s a 50 MHz gain accompanied by something else, say, an upgrade from a K6-2/400 to a K6-III/450, in which case you’d get a larger gain, maybe 25-35 percent, due to on-chip cache. But with CPU speed being a fairly small factor in overall system performance, that 35% increase definitely won’t work miracles.
When upgrading a system, I generally attack system RAM and the hard drive first. It’s amazing what a difference dropping in a 7200-rpm hard drive makes. I recently made a P200 boot Win95 in 15 seconds by replacing the drive, dropping in another 64 megs of RAM, then doing a fresh Windows installation and tweaking msdos.sys. Very nice.
Good observation. Thanks.