Once a plausible rumor appears both on Reality and Mac OS Rumors, that's about as official as it gets without Apple confirming it themselves. So at this point, we're totally willing to accept that the iMac will indeed ship with a software-based modem. Depending on just how much speed Apple's engineers can squeeze out of their modem software, the wisdom of that decision ranges from "poor technical decision, but excellent cost-savings and marketing move" all the way up to "excellent decision for everyone."

The up sides to the software modem decision are obvious. First of all, it's free, once you account for the development costs, so it will add essentially zero to the iMac's prduction costs and you know Apple won't get itself stuck in one of its classic Three Stooges scenarios where thousands of people are lined up with money in hand, but Apple can't build any more iMacs because they're out of modems. Secondly, it means Apple can upgrade its users' modems simply by releasing a software update-- and according to Rumors, Apple plans to do just that, to bring early U.S. iMac buyers from 33.6 up to 56K at about the time of Mac OS 8.5's release. Apple can therefore start hawking the iMac as having a 33.6 modem "with a free upgrade to 56K," which should alleviate at least some consumer concerns.

The only serious potential down side is performance. The rumors are that Apple has completely rewritten its Geoport Telecom software modem code to use much less CPU power, and with full G3 optimization, so that it won't use more than about 5% of the CPU time when running. Even so, we at AtAT remain skeptical, since a 33.6 Geoport connection yields about half the throughput speed of a 28.8 hardware connection on our PowerTower Pro. It'll be interesting to see just how fast Apple can make the iMac's modem move.

Do you know anyone who'll buy any piece of junk they come across that has, say, the South Park kids emblazoned on it? (Not that we aren't fans of South Park; it's a random example.) Because that appears to be the mentality that analysts expect among early Windows 98 buyers; some people will buy any operating system with a Microsoft logo on it. There's some very interesting analysis over in a Reuters article on Yahoo! News.

Chris Le Tocq, an analyst at Dataquest, calls Windows 98 nothing more than "a packaging exercise" which is completely unnecessary for Windows 95 users who have a working system. But despite that assessment, and in spite of the lukewarm reviews we've seen of the product so far, he still expects over five million people to buy the $90 upgrade this year, and another eleven million to shell out the cash next year. Do the math, and you can very quickly see why Bill Gates is the richest man in the world (and climbing). The man sits at the head of an empire that can rake in obscene amounts of cash by sticking a label on a mediocre product, because it'll sell anyway.

This is not to say that we Mac folk don't follow the same instinct-- heck, slap a six-color Apple logo on a week-old baguette and wave it around at a users group meeting, and you can bet someone will fork over some cash for it. We just find it interesting that an analyst could say something to the effect of, "Windows 98 brings essentially nothing new to the party. But it'll generate $1.3 billion in revenue anyway." The point is, Microsoft is huge enough and has an installed base wide enough to generate well over a billion dollars just by packaging a collection of bug fixes and a web browser. Now that's mighty impressive.

It seems like the only things that happen in the tech world anymore are litigation and buyouts; everyone's suing everyone else, and the monotony is only broken when one of the players swallows another one whole. Not that that's necessarily a bad thing. Take CompUSA's announced acquisition of the Computer City retail chain, for example; CompUSA sauntered up to Tandy Corp. and said, "Gee. it's a shame Computer City's losing you money. Y'know, we'd take it off your hands for, say, $275 mil." And so the deal was done; read the details over at ZDNet.

Why is this a good thing? Well, it certainly isn't good for everyone; for one thing, if CompUSA closes many of the Computer City stores, that's one fewer place where consumers can buy their equipment, and less competition, which we never like. But it could be a good thing for the Mac universe; remember, Apple yanked all the Macs from Computer City months ago, when it became apparent that it was just another retail chain that didn't care about selling Macs. Now that CompUSA is the only national retail chain that sells Macs, it behooves us all to see CompUSA grow and thrive. (Not that CompUSA's doing such a hot job of displaying and selling Macs themselves these days, but we still have high hopes.)

Whether or not any of the 100-odd Computer City stores will be converted into CompUSA's remains to be seen. We're guessing that just about everywhere there's a Computer City, though, a CompUSA isn't too far away. Apple may get a tiny boost from this buyout, since no more Computer City stores means more buyers will go to CompUSA instead-- where at least there's a non-zero chance that a first-time buyer will get a Mac instead of a Wintel, especially once the iMacs are out.