Subject: Re: More on PIII > LOL -- trouble *starting* to write? that's me.. information communism in its purest form: from each according to their inability to shut up, to each according to their boredom threshold. ;-) ('nother 14 on the Beaufort scale, BTW.. select your provisions accordingly) > one of your points (the myriad of choices that must be > made) is valid. The problem, in my mind, is that MhZ has become > a surrogate for "good" -- ie, the faster, the better. I think > this has to do in large part w/Intel's marketing campaign, but > it's borne out here, in a c|Net article on the iMac: i've seen similar articles elsewhere, and variations on the same theme at all levels of fervor in the smouldering brushfire which is online discussion of the iMac. without marching into that particular inferno, i agree that computer marketing has latched onto clock speed, and waves it as a battle flag wherever possible. this is actually an issue i've been following for more than a year.. i tend to follow news to watch the evolution of spin, rather than for the atomic facts in any given story. a year ago, the standard interpretation of the MS/DOJ trial was that the government was grasping at straws, for instance. now, it's more or less "here's how Microsoft shot themselves in the head today". the pre-release discussion of the iMac averaged somewhere around "last gasp of a dying company, trying to fob off another inferior product with a pretty (sic) box". by now, the press and pundits have decided that the iMac is somewhat neat, if your tastes happen to run that direction. the issue of Intel's marketing of processor speed has also changed flavor over the last year. the fact that a G3 PPC beat Pentia with higher clock speeds on Intel's favorite benchmark forced Intel to spend several months fighting the inertia of their own "yes, size *is* everything" message. the fact that Apple was able to co-opt Intel's own marketing symbol with the infamous 'Toasted Bunny' commercial was a coup. the 'Snail' and 'Steamroller' commercials were very loud challenges which got very quietly ignored.. they weren't victories in and of themselves, but they opened some cracks in what was previously a solid position. clock speed is one of the few simple numbers associated with computers, so it will never entirely disappear from the marketing. OTOH, the rise of Linux and the resurgence of Apple are laying the groundwork for some serious paradigm shifts in the future. (okay, i apologize to all those who hate the 'p' word.. i despise it myself, but that's what we'll see it called, so we may as well start building up a tolerance for it now) the overall uniformity of the Wintel dominance of the desktop has allowed MS and Intel to push certain basic assumptions into the general consciousness, one of which is "more megahertz is good". (and a second apology to lovers of the english language for using a pseudo-plural in a collective, but singular, context. grammatically correct, yes.. euphonious, no) as long as you're working with the same OS and basic set of applications, that idea is largely true.. most of the parameters involved in the benchmark are being held stable, so you can see the effect of a single change. Apple and Linux challenge that mindset by changing different variables. any Linux junkie will tell you that switching the OS gives you significant performance gains on the same hardware. therefore, Windows starts looking like a drag factor relative to the CPU. users who run Photoshop first on an iMac, and then on a 233 MHz Pentium, are also likely to see differences in performance. some of that difference can be attributed to the OS, but most users will see it as a weakness of the hardware. right now, the population of dissenting users is small enough that they don't represent a serious threat to the mental status quo. the critical point will come when about 15% of the population owns or works with at least one non-Wintel machine on a regular basis. i won't bore you with the math, but it's basically an evolution thing.. most people require an average of three exposures to a principle before they recognize it as such. the first time, it's just random noise; the second time, it's deja-vu; the third time, it's a pattern. if 90% of your contacts support one mindset, they overwhelm and submerge the remaining 10%. instead of going from deja-vu to seeing a pattern, you go from deja-vu to calling it a fluke. instead of meeting another person who says "yeah, i've always thought Photoshop was faster on my iMac", you'll meet someone who tells you it's a side effect of your video card, or that you need more RAM, or something of that nature. OTOH, when 15% of a population are walking around at the deja-vu stage, the odds of meeting someone else who can help you close the pattern hit a break-even point. instead of being drowned out in random noise, the pattern becomes self-sustaining and begins to compete openly with the established pattern. the established pattern can no longer maintain dominance through intertia, and life becomes interesting. Linux has made an impressive splash because unix achieved more than 15% representation on the internet long ago. a majority population was willing to accept this particular variant, so the Linux movement became self-sustaining online. basic acceptance of Linux has become a dominant pattern online, and the online population makes up more than 15% of the population as a whole in most technological nations. by achieving dominance in an isolated environment, Linux has been able to leverage a mindshare of about 2-1/4% into a general awareness. meanwhile, the iMac bodes fair to bring Apple up to a 15% mindshare all by itself. the fact that it's visually distinctive creates another type of isolated environment.. one which serves the same purpose as clock speed, but uses a different angle of approach. clock speed is simply a way to distinguish between boxes that look the same, and the iMac breaks that distinction by offering another. the iMac will compete with the lowest end of the PC market, precisely where the odds are highest that vendors will take shortcuts. there will be strong temptation for cheap and unreliable sub-K PCs to cash in on the reputation for quality which has been earned by more expensive machines. the beige box market will become very noisy, but the iMac will stand clear because its performance can be associated with its appearance. IMO, it will also dominate in terms of quality. whether you like the case or not, most people will admit that the iMac represents a very carefully crafted example of human-interface driven design. the noise among biege boxes will have secondary effects on Apple's market share in the high-end market. once again, Apple's products will be visually distinctive, which isolates them from the noise, and ties them to a dominant reputation for quality earned by the iMac. meanwhile, high-end PCs from other vendors will suffer guilt by asociation with the worst profiteers of the low-end market. even if PC manufacturers start trying to build their own visual brands, their success will be limited thanks to the "get components anywhere" mindset which has been pushed as another good in the general Wintel world view. (that's not a winning strategy unless you have a monopoly on a vital component, BTW.. it was that weakness which forced IBM to hand off so much power to Microsoft. IBM lost control of the BIOS, and Microsoft is at risk of losing control of the OS) at that point, buyers in any market will have a decent chance of getting firsthand information about performance comparisons between Macs, Windows machines, and Linux boxes all running at the same clock speed. with any luck, there will also be a wide enough distribution of PPC Linux boxes to show the comparison between Linux and the Mac OS on roughly equivalent hardware. at *that* point, clock speed will start losing its relevance in the general mind.. other issues, like "what do you want to do with it?" will start to become more important. there will still be plenty of people who fall back to the old argument of clock speed, but there will also be plenty on hand to say, "yeah, right."