Cringely: OS X on Intel

sti writes: "Cringely's column this week argues that Apple should port OS X to the Intel platform. He makes an interesting case for it. I would definitely favour this. I've always had this warm spot in my heart for Apple but rarely had the money to pay for their overpriced hardware."

Comments

Wouldn't be the same

[Jebediah21]

OS X on Intel just wouldn't have the same experience. When you buy an Apple machine you know that the OS is well tuned to run on that hardware. You don't have to worry about an odd mix of hardware or bios problems that are responsible for a number of woes on x86.

I think the only way for OS X to be viable on x86 is with different pricing. Say something like $50 for no support, but $150 with support. That way way nerds like us can play around with a leet OS cheaply, while those who need support would make up for lost hardware profits.

Re:Wouldn't be the same

[Stenpas]

MacOS X has a very far way to go in optimization.

You'll see a huge leap in performance after they get aqua accelerated via the graphics hardware. Since the graphics hardware (which these days is insanely fast) will be handling it, the CPU will have more power to use for other things. How much more? I don't know, but judging from the looks of things (anti-aliased, alpha layered, bezier curved, quartz rendered, drop shadowed, etc), I'm sure it will be significant.

Window buffering isn't turned on by default, so a 800x600 window at millions of colors eats up 1.9 megs of ram. If you're the type of person who likes to have 70 windows open at a time, this adds up very fast. With window buffering, each window will use 8.5-10x less memory. So with those 70 windows, instead of using 133 megs of ram, they use 15 megs. That's a lot of ram that could be going elsewhere, and since you won't be using as much swap or any at all, you get a huge speed increase.

A big one which can't be dealt with on a technological perspective is our dependancy of the Classic compatibility environment. Some people like having it open at all times for maximum compatibility. Well, even if they don't, having to open that One Small Thing(tm) in Classic is a pain in the ass because it uses an astronomical amount of CPU power and Ram. So the sooner we lose this dependancy, the better.

And from the looks of it, getting MacOS X synced up to FreeBSD 4.5 might be good. I'm sure we all love "hundreds of fixes, updated many system components, made several substantial performance improvements, and addressed a wide variety of security issues." Enough said.

After it's all said and done, I'd at least hope that it would be on par with MacOS 9. A little slower, yes, but not drastically.

So when is all this coming? It would have to be on or around March 24th, 2002. That's when the transition to MacOS X is supposed to be complete. What better way to celebrate than a major upgrade? If we're still bitching about something as general as speed after the transition is all said and done, then either Apple failed with MacOS X, or they need to extend the transition period.

But Why?

[countach]

Cringley himself answers his own stupid question... Who would buy such a beast? Mac users buy Mac hardware, so why bother? That's exactly right Cringley, so the product would be a waste of time. Either the Mac users would save some pennies on Intel hardware and Apple loses, or they wouldn't
and it would be a waste of time. Most users are simply not going to bother loading another OS with Windows, that's why BeOS failed. Linux is making some headway because (a) it's free as in beer and (b) it's free as in liberty. We don't need another stinkin proprietry OS, one is enough and users know it.

Makes Apple a threat

[murphro]

Cringely points out possible benefits to Apple if they enter the OS market on Intel (and has several good points). But what about the certain negatives? Apple now is a mild threat to MS's power. But if they 'infringed' on turf that was MS's, they would certainly be targeted by the giant. Is it really in Apple's best interests to rouse that big of bully? I don't think so.

Cringely mentions Netscape in his article (how by competition, MS made IE better). Look what happened in that case. Would Apple want to risk the same fate? To sacrifice themselves so that Windoz might be a little nicer to use.

Come on.

Crap

[gargle]

The thrust of Cringely's argument, which he devotes most of his article to, is this: Apple should port OS X to Intel because "it is exactly the competitor Microsoft needs." But what really matters to Apple is: Will porting OS X to Intel make Apple more or less profitable?

Cringely resolves this complex matter in the space of a paragraph length assertion "The upside for Apple is enormous. Suddenly, their software budget is leveraged across a much larger number of units, making the company more profitable and able to spend even more on making the software better."

Really, Cringely? I think we need more than a handwaving assertion to back this up. e.g. What effect will porting OS X to Intel have on Apple Hardware sales? What will MS's response will be - will it withdraw its Office and IE products for OS X? etc.

Re:Crap

[gwernol]

Cringely resolves this complex matter in the space of a paragraph length assertion "The upside for Apple is enormous. Suddenly, their software budget is leveraged across a much larger number of units, making the company more profitable and able to spend even more on making the software better."

Actually I think he's right about profitability. Apple typically makes a much greater profit per unit of software than per unit of hardware. There have been years when Apple's entire profit margin has been from its software division(s).

The problem that Cringley misses is that Apple has to think not only of its profits but also of its revenues. If it lost the hardware business it would immediately drop its revenues from $8 billion to around $500 million. Even if its profits went up at the same time (which they might), they would get crucified on Wall Street for this. No sane company would ever pursue a strategy that involved such a dramatic cut in its revenue stream.

So even though Cringely is right about profitability he ignores the revenue impact so his overall argument is flawed.

We wish.

[G-funk]

It's simple. Let's say apple release OSX on intel. Forget their hardware sales, forget support problems. This would be the future:

1) Office is no longer available on any apple lines, neither is Explorer.
2) Office XP++ doesn't write in any format office X can read.
3) Office was never available for OSX on intel.
4) Microsoft tells Dell, HP, etc that if they want to offer OSX then windows wil cost $$$$ more per copy.

which leaves apple going steadily bankrupt, and the masses with no options if they want user-friendly but don't want Bill....

I'd love it, I'd be first in line to buy it, but it ain't gonna happen

Re:Its going to be hard

[GMontag451]

Everyone is saying that supporting all the hardware is going to be a real chore, but I disagree. I think the Darwin people will be able to take care of that. Darwin has already been ported to x86, and I think it supports a rather large set of hardware. All the hardware support, except for maybe video card specific graphics acceleration, would be done at that level anyway. Another option would be for Apple to sanction a line of existing or yet to be made x86 hardware that is MacOS compatible.

The real problem porting would be Quartz. From what I understand, Quartz is rather heavily optimized for AltiVec. They might be able to help the x86 version along with better video card acceleration, but they would probably have to settle for slower speeds there anyway. The other problem would be Classic. If Apple even bothered porting Classic to x86, it would run incredibly slow because it would have to emulate a PowerPC as well as a 68K chip.

However, a port to x86 would bring up some very interesting possibilities, such as a WINE type system for running Windows binaries, rather than a Virtual PC type full emulation. Or perhaps an end to this stigma MacOS has in the eyes of game developers.

Cringley missed something.

[_typo]

What we need isn't Mac OS X for Intel. What we need are cheap PPC machines, with dull beige designs.

That way dual-booting might actually be a nice thing. On one side you have linux, on the other you have OSX, a beautiful and powerfull OS, not some Microsoft piece of crap. Plus we get nice hardware. Altivec anyone?