Wintel Easy PC – the substance-free initiative?
Microsoft seems to have hijacked Intel's cunning plan -- is it clever, or just desperate?
Analysis Here's a puzzle. Microsoft didn't exactly announce much of significance or show much that was convincing at last week's WinHEC, but if you try to look for further information on what was potentially one of the biggies, Easy PC, you'll come up strangely empty. Now, why could this be? Microsoft itself doesn't go much beyond the press release, which is headed "Industry support grows for Intel and Microsoft Easy PC initiative." This is described as a "multiyear vision... targeted at improving the overall experience for PC users." A clutch of major PC manufacturers has toed the line, and according to Microsoft will be developing new PC prototypes by the end of the year. Ah, but what's going to be in them? Our own dear Graham Lea suspects a dastardly plot to eliminate ISA slots, but actually, on the dastardism scale this probably doesn't register very high, and anyway it's an old idea that makes sense. Intel and Microsoft have been trying to roadmap ISA out of the picture for several years now, and one day they're going to succeed. Intel has been particularly active here, and is pushing a "Legacy Removal Roadmap" that starts with the elimination of ISA devices and ISA slots in the second half of 99, and then moves on through the progressive removal of all sorts of other stuff -- gameports/MIDI, PS/2, serial, parallel, IDE, floppy, VGA and user-accessible slots. In Intel's view, getting rid of the old stuff is the first step towards making PCs easier to use, and of course old Satan is absolutely right here. If we substitute smarter plug and play technologies for the old confusing stuff life will be a lot easier for all concerned. So Intel is keen on USB and stuff like Instantly Available/On Now. Its enthusiasm for IEEE 1394 may be a little muted at the moment, but as far as we can gather this is something to do with a campaign to price-gouge on licence fees. Note however that these are very Intel concerns, and that Intel is currently happily describing the Easy PC Initiative as "previously known as the PC Ease of Use Initiative," as announced at the Intel Developer Forum last autumn. A reality-check of the original announcement confirms that there was no mention of Microsoft participation, and at Intel's showing of a "Concept PC" demonstrating ease of use at Comdex in November, there was again no obvious Microsoft participation. Intel's Easy PC announcement last week describes the Initiative as a "joint Intel/Microsoft" one, but the evidence pointing to it being one Microsoft has muscled in on fairly recently seems pretty conclusive. According to Microsoft, Easy PC development will be focussed on four main areas: Easy set-up; Easy expansion; Ease of use and New form factors. The detail of these is worth examining. Easy expansion obviously just continues the Intel campaign on legacy removal and migration to USB and (maybe) 1394, while new form factor just means new Intel motherboard designs, FlexATX in this case. Ease of use is more of a joint effort. The addition of OnNow is of course an Intel component, as is the "reduction of legacy hardware." Microsoft contributions here will be "faster operation with improved, task-based user interface" and "self-repairing features" -- here be dragons? Self-repairing software is one of Microsoft's hot buttons, but probably isn't that do-able (maybe we should write software that isn't self-breaking first), while the improved user interface sounds like something that will be attached to one of those new operating systems Microsoft has so much trouble finishing. But look on the bright side -- this is a "multiyear initiative," so we don't have to ship all the components right away. Easy set-up is also one that lies largely on Microsoft's turf, but it'll probably turn out to be more to do with horse-trading between Microsoft and the PC manufacturers over who gets to do what in the initial set-up procedure than it is to do with breakthrough technology. Microsoft does however suggest that machines could be "preconfigured for connectivity," and we could interpret that as meaning that PC manufacturers are going to be allowed to do a little bit more of their own thing. But shall we get back to the roadmap? If there are going to be prototypes by the end of the year, and machines based on it will be in the shops next year (will have to be, if ISA is going to be dead by then). Microsoft has a "new" OS, Windows 98 Second Edition, due shortly, and is planning a consumer-oriented follow-up for 2000. It can't do much in time for Second Edition (this "will improve the initial out-of-box experience for user," says Microsoft), and considering the timescale, it will be difficult to do much radical in time for the 2000 version. Some more hardware-oriented Intel initiatives could be supported by Q2-Q3 (and maybe supported earlier by refreshes of the OEM version), but trickier stuff like an "improved task-based user interface" sounds a lot like the sort of technology that would knock shipment back into 2001 or beyond. So actually what we've got with Easy PC turns out to be largely hardware-related stuff that was in the PC99 design guide, and that has been pushed further by Intel independently since then. Microsoft input is a lot vaguer, and probably a lot further away too. But we have a theory. Note first of all that Intel refers to "new features for Windows 2000" but doesn't mention future versions of 98. This isn't particularly surprising, as absolute confirmation that there would be new versions of 98 wasn't forthcoming until WinHEC last week. Note also that the sole Easy PC 'innovation' Microsoft is talking about for Second Edition is something it was going to do anyway - 'improve' the out-of-box experience. And then note that Microsoft has decided to cut its losses on the Windows 2000 consumer version (at least for the moment) and substitute a series of simple, easy-to-ship upgrades to Windows 98 instead. What could it put in them? Well, if it took on board some of the work Intel had been doing and called it a new initiative, rather than just PC99 mark two, wouldn't that do? The $64,000 question, of course, is how long Microsoft can confine itself to shipping the easy stuff before it gets tempted back into three year OS development cycle. ®