Slashdot Mirror
OS X on Intel Hardware?
CNN has an article originally from Macworld on the possibility that OS X will be ported to Intel-type hardware. There is much mention of Wilfredo Sanchez and his recent port of Darwin to Intel compatible hardware, and the economic state of Apple and how that may shape the decision to run with this idea. Most of it is speculation, but definitely interesting if it comes to pass, seeing as not much would be needed in the porting effort from where they are already.
The database seems to have got corrupted, and various articles are only appearing in the sections, rather than as part of the main story flow. You can't even get to this one (or the NT Hosting Trollfest) via the Previous and Next article links below the articles.
You're right though, this IS important news!
Anyone remember when Power Computing used to kick their ass performance-wise in the PowerPC realm?
This was largely an illusion due to volume differences. Power Computing sold maybe 200,000 machines in their entire lifespan. Apple seels about 1 million per quarter. When you have such low volume as Power Computing, you can announce products and ship them much sooner.
- Scott
------
Scott Stevenson
Scott Stevenson
Tree House Ideas
This "OSX on x86" seems to show up almost as often as Apple being purchased by Sun.
I don't think anyone has a religious attachment to the assortment of microelectronics that make up the PowerPC chip. It's just a chip, after all. The issue is that there are some fundamental hurdles to overcome if Apple was to release an OSX on x86 and not have the shareholders lynch Jobs. None of which I've seen the osxonintel.com people address, by the way:
1) Profits: How does Apple succeed by competing directly with Microsoft and Linux on the same hardware? How do they differentiate? Much of Apple's core value to its customers is because of the tight integration of the hardware and software.
2) User experience: How does Apple maintain the Mac ease-of-use with such a plethora of x86 hardware to support?
3) Applications: All of the legacy Mac apps (Carbon and Classic) are processor-dependent. A great OS is useless without applications.
4) Product introductions: Apple is currently able to rapidly introduce new technologies in both the hardware and software simultaneously to get products to market quickly. These include FireWire, Airport and many other low-level improvements that don't have sexy names. Waiting for hardware manufacturers to get their act together would hurt Apple.
Some of this may be solved by Apple creating a proprietary machine, much like a current Mac, but with an x86 chip instead of PowerPC. That may solve all the problems except #3. How do you run current Mac apps, especially those that are so sensitive to performance like Final Cut Pro? I know all Slashdotters think that a Pentium III 1.2GHz is worlds faster than a G4/500, but in real-world use, this just doesn't seem to be the case. This seems to be especially true for media-intensive applications that the G4 was designed for. And there's only so fast Office is going to launch. Regardless of which chip is faster, it's clear that the x86 is not a significantly faster chip to emulate a G4 in any sort of reasonable manner. The only solution, therefore, would be for Apple to build some kind of hybrid machine with both a G4 and Pentium/Athlon on board.
And then there's the most basic question: why? Yes, Intel has continually ramped up the megahertz of the Pentium (as has AMD), but the G4 has held its own in real-world performance. The biggest issue to tackle is the perception problem that Motorola has created by not upping the clock rate of the G4 in about a year. But Apple has addressed this somewhat by shipping MP systems at the same price as single CPU machines. And while this power is not really exploited by Mac OS 9, Mac OS X (the topic of conversation) was built for SMP. In that case, you can get a system with two fully-utilized G4s on board for $2500, the need for an x86 system is greatly deemphasized.
Finally, there is this weak but persistant argument that as soon as Apple releases Mac OS X for x86, their marketshare will explode because everyone already has the hardware. I flat out just don't buy that. Sure, maybe hobbyists would pick it up, but with no apps (Office and IE, for example), it's useless for real work.
- Scott
------
Scott Stevenson
Scott Stevenson
Tree House Ideas
Nothing blazingly new in the article, but having cnn coverage of that is nice.
Note that Carbon is not supposed to work on i386 now (while Darwin and the Cocoa environment does). Quartz (the window manager) either. So it may be a little more trouble than the ports that NeXT did.
Anyway, the real killer that apple have is the Yellow Box. It is the Cocoa environment on Windows, and this would be a very very smart move to make it avalaible again (in that case, people could write windows apps in Cocoa). Without that, the developer mindshare of apple is going to shrink more and more (because hard-core mac developers have little reasons to move to Cocoa: they would loose Mac OS 9 compatibility, and have to learn a new language. And the less Cocoa developers, means less Cocoa apps, means less possibilities to move to another processor/OS, which in turn means less appeal for the mac platform, which means less developers....)
Cheers,
--fred
1 reply beneath your current threshold.
They have a reasonably strong OS, they have a lead in multimedia software, yet they have refused to acknowledge plans to capitalise on the huge x86 OS market. I don't get them.
Exactly. You do not understand their core customer base. The Mac users would not be tolerant of the hardware nightmares that are associated with x86 hardware. Not to mention losing all Mac applications (they are processor dependent).
Just because it's they way you personally think things should be, doesn't mean it the best thing for the platform, the company or the userbase.
Nobody gets rich selling just hardware.
Sun seems to do okay.
- Scott
------
Scott Stevenson
Scott Stevenson
Tree House Ideas