Cathode Tan - Games, Media and Geek Stuff
logo design by man bytes blog

Friday, June 10, 2005

A Side of Apple

OK, having now had time to collect my thoughts ... and read the analysis of much smarter people, this is what I'm guessing with Apple's plans with Intel.

No Apple WINE
The more I think about my comment that WINE might help kill Longhorn, the more I realized how wrong it was. History has proven that you can't beat a platform by simply emulating that platform. It's handy to have - and I think some company like Transgaming will pay attention to it - but it won't be Apple.

And maybe they don't need to...

Mostly, it's about compatibility
Apptels would be cheaper, but possibly not by much. Without the benefits of a massive clone hardware framework, and considering Apple's small market, and that Apple's desire to have OSX be Mac only will require custom mobos, I wouldn't expect it to go down much. We'll get more powerful computers for about the same buck we're spending now, I'd think.

Also, I think the optimism that in a few months, someone will hack OSX to any Intel box may be optimistic. Times have changed for motherboards and there is a lot more interaction these days. I had to call Microsoft just to reinstall Windows onto a new box for god's sake. Not saying it definately won't happen, but I wouldn't assume it's inevitable.

But the real boon is bringing the Apple software under the same architecture as the Windows world. It removes a serious layer when it comes to writing cross-platform apps. For gaming, this could be major. I agree that the current port companies are probably in trouble - they won't be able to charge an arm and a leg anymore to port a single title. In their place, I think, we'll see more portable middleware and companies which handle more small jobs. If Apple allows it, some applications might be Intel only. If they did allow that, it might pave the way for publishers to release Mac and Windows versions of their games on a single DVD. With an Intel based OSX, publishers would have a cheaper solution than porting to PPC and a larger user base than Linux - making the costs of a port much more reasonable.

Intel only, you ask?

Rosetta may be Dumbo's feather
But let's face it, that doesn't sound nearly as good. However, having software translate code from one architecture layer to another sure sounds like good science fiction - but one shouldn't expect too much from it in reality. I would think of Rosetta as more of a crutch than a miracle. And I'm prepared to lump the Universal Binary into that. Apple's done this crazy stuff before - with the introduction of OS X. They know how to make enormous transitions work, and this isn't as drastic as going from Classic to Carbon.

But I don't think anyone should fool themselves into thinking that the goal is anything short of native Macintosh apps running on Intel iron. Apple will support their PPC brethren only as much as they need to in order to avoid stabbing their loyalists in the back. And that's not to say that efforts like Rosetta aren't impressives steps for supporting the user ... but we should view it for the temporary security blanket that it is.

Think small
Apple won big with the mini, and Intel's hardware is much more impressive at running small than IBM. One thing that has been oddly missing from Apple's bag of tricks has been a more mobile edition of their OS. Windows has it. Linux has it. Why can't I run OSX on a PDA, anyway? Granted, handhelds don't use the x86 framework. But they do use Intel.

At the very least...

We're not done yet
The whole announcement was odd. No specifics on the chips, no mention of anything rhyming with "entium". Cringely asks a valid question - why not AMD? Why is the developer box so slow? The whole thing looks like Jobs has his fist clenched, ready for a follow up left hook. Partnering with Intel for some serious custom chipwork? Apple's always been fun to watch for announcements, and this should keep us entertained for a while.


Corvus said...

Great analysis, Josh. It'll be interesting to see how Apple goes about locking the OS to their hardware.

Tony said...

However, having software translate code from one architecture layer to another sure sounds like good science fiction - but one shouldn't expect too much from it in reality.

Isn't this what Transmeta tried to do with their Crusoe processors? We can see how well that turned out for them...