% fortune -ae paul murphy

Cringley, MacTel, and nutty theories

When Apple first announced its MacTel decision Robert Cringley started to spend a lot of time looking for the long term smart in Apple's otherwise inexplicable decision. As a result he first used his PBS pulpit to announce what I thought was a fairly nutty conspiracy theory in which the whole thing was put up so Intel could cash out Steve Jobs and then go after Microsoft for world PC domination.

Later he dropped that idea to look for a Jobean second shoe in Apple's predicted purchase of Adobe:

Over the past three weeks, we've laid out in this column a sequence of clues and events that suggest Apple is planning to next year take on not only Microsoft's hardware OEMs, but also possibly Microsoft, itself, by leveraging a vestigial legal right to some portion of the Windows API -- in this case, literally the Windows XP API. This bold strategy is based on the high probability that -- if something called Windows Vista ships at all next January -- it will really be Windows XP SP4 with a new name. Microsoft is so bloated and paralyzed that this could happen, but what's missing is an Apple application strategy to go with this operating system strategy, because Microsoft's true power lies not in Windows, but in Microsoft Office. Fortunately for Apple, I believe there is an application plan in the works, and I will describe it here.

...

Office is how Microsoft makes most of its revenue, and Office is the bludgeon Microsoft uses to keep other software vendors in line. Without Office, Microsoft is just a company with an archaic and insecure OS. If Apple does go ahead to compete head-to-head with Microsoft for Microsoft's own Windows customers, Cupertino will have to be ready in case Mac Office is withdrawn and Windows Office mysteriously breaks on Apple hardware. There is a good likelihood this won't happen, especially if Microsoft can find a way to rev Mac Office for IntelMacs sorta running Windows -- a hybrid product that would look better than the Windows version while retaining 100 percent compatibility and generating an enormous new revenue stream for Redmond. This is the carrot Apple will use to keep Microsoft from doing something truly destructive

...

Steve wants Windows applications to run like crazy on his hybrid platform but to look like crap. In his heart of hearts, he'd still like to beat Microsoft on the merits, not just by leveraging some clever loophole. So he needs the top ISVs who are currently writing for OS X to continue writing for OS X, and that especially means Adobe.

There's only one way to make that happen for sure, and that's for Apple to buy Adobe

A bit later Cringley got into a mud slinging match with John Dvorak over the inner meaning of the bootcamp option allowing MacTel machines to boot and run Windows/XP. Here's part of one of his comments on this:

So Apple will at least offer the option for users to run a virtualised version of Windows Vista atop OS X, which brings with it two HUGE advantages. First, the bad guys and script kiddies will have to get through OS X security before they even have a chance at cracking Vista security. Second, by running a virtual version of Windows Vista loaded from a read-only partition, Microsoft's recommended method of dealing with malware (periodically wipe the OS and application from your disk and load them anew) can be done in seconds instead of hours and can be done daily instead of monthly or quarterly or yearly.

So far Cringley's track record on Apple hasn't been good, but this is one prediction that I think may have merit because what he's really talking about is using Apple's 1997 patent exchange agreement with Microsoft to allow Apple to update and use Sun's WABI code for the Windows/XP APIs - thereby enabling Windows applications to run without Windows licensing or Windows overheads.

WABI/XP is technically do-able, and the original WABI certainly ran Windows 3.11 applications faster on both HyperSPARC and PA-RISC than they ran on Intel, so this could happen.

But please notice two important ancillary issues here:

  1. the other technical strategy for making Windows on Darwin work well is now called virtualization but goes back to running CPM86 under Microsoft Xenix before Gates contracted for QDOS and got into his deal with IBM. Although Windows software bloat killed this in the late eighties, recent decreases in memory costs and the development of more capable memory controllers for x86 have brought it back at the OS level. Thus it's possible that Apple's product plan includes more of this, perhaps at the application plus libs level rather than the OS level.

  2. Sun currently has a patent cross licensing deal in place with Microsoft - and, unless the agreement has an explicit limitation on this, could therefore bring WABI/XP to market for its own desktops, whether workstation or Sun Ray.

Fundamentally, however, both of Cringley's more probable hypotheses to date: WABI/XP and an Adobe mediated Apple Office, have the same flaw as explanations for Apple's MacTel decision: both strategies would have been better supported by decisions to stay with the PowerPC or switch to SPARC than by a decision to use the same processor Microsoft programs for.

So what really happened? Personally I think Steve Jobs had grown increasingly unhappy with IBM's claimed inability to meet its performance promises for Apple while exceeding them for its own use, for Microsoft, and as part of the Cell partnership; saw IBM's control of Cell as a honey pot from which Apple wouldn't be able to escape; thought he had a good "plan B" in place because his software people had kept the Intel option current since the original G5 decision; and therefore pulled the plug on the IBM relationship in response to some IBM action - possibly another production delay for the low power G5 - only to discover after the fact that the industry had long since left Plan B behind.

Tomorrow, however, I want to compete head to head with Cringley on the development of a better conspiracy theory aimed at rationalising Apple's decision - stay tuned, all bragging aside, my record on these things is a lot better than his.


Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.