Mac on Intel
I've just got to throw my two-cents in here. Well, penny-and-a-half, anyway. I'm pretty good at digesting industry news and translating it into verbiage normal people can understand.
Recently Apple announced that it would be redesigning and selling Macs using Intel processors instead of the PowerPC chips they use now. This has led to much buzz and speculation throughout the industry as well as in my peer circles.
In the professional circles, many ignorantly dismiss the Mac. Whatever their reasons, the Mac is beneath them or theirs. Most of those quick to discount them will eventually admit to having no relevant experience on a Mac; some perhaps limited to clicking a few icons on a Mac they encountered at a local retailer, and more than that limited to TV commercials or less!
In less-than-professional circles, many dismiss the Mac on a more realistic front; the particular software need I have is unavailable. "The game I want to play," is what that most likely means. While the software selection for the Mac is wide and nearly complete when compared to software available on Windows, there are gaps and holes. Even the "same" software (i.e., Microsoft Office) is not truly the same.
Now, I'm no Mac expert, but I do pride myself on being "platform independent." In addition to experience with Microsoft environments since the late 1980's, I've long played with the UNIX and UNIX-like variants--since I was introduced FreeBSD in the early 1990's. On that trend, I run FreeBSD, LINUX, Solaris, and Mac variants in my lab at home.
Woah! Why's the Mac in that list?
Well, under the hood of Mac OS X is Darwin, a UNIX variant based on BSD and NeXT. On top of that are the things that make us think "Mac," stuff like the interface (Carbon) and the applications (e.g., iTunes). The interface is more key to our experience, as the applications, outside of the interface, typically don't care about hardware.
Given this, I often counter, the Mac is not the oddball the pundits assume it to be, based on its previous, separate, and proprietary operations. Current LINUX open-source projects generally easily compile and install on the Mac, giving us, therefore, very familiar tools, applications, and environments in which to work. They look a little different around the edges (literally), but function the same on the inside; you type on the keyboard, and the letters appear here...
Now that we've got familiar tools and applications, it should be easier to port our software to the Mac. The Mac Xcode developer tools make much of this transition less painful. There are even tools that portend to directly translate Win32 code to MacOS. Shell script, daemon, X, Java, and the like take very little, if any, translation.
I've digressed, but just a little. My intent with the digression is that Darwin, the core to Mac, is available for Intel processors now. You don't get the Apple look and feel, but that'll come soon, I'm sure.
So far I'm not focusing on the Apple-to-Intel shift. Let me get back to that.
The shift doesn't mean one thing that many I've talked to assumes; that like LINUX and Solaris, Mac will become another platform for the same PC hardware that runs Windows now. It might work out to be the case, but it currently hasn't been pledged that you'll be able to buy a PC running Mac from Dell...or take your current Dell and install a Mac OS on it. Replace Dell with your personal favorite (I generally build my own).
One thing to remember is that Intel makes more chips than the x86 line (which includes the Pentium). One article I've read suggests that one of the Intel RISC processors is a more likely candidate than the x86 line. I've been trying to find information on the Intel-based Macs that Apple is currently leasing to developers; not so I can get one, but just so I can learn what to expect.
Whether they use the x86 or another line, I think Apple will still build their own mainboards and hardware; choosing their own components, not taking the varied and occasionally weird combinations that are on your favorite vendor's shelves and websites. The control they earn by having their boot ROM and PCI bus provides opportunities over the varied stuff we put up with on our typical x86 mobos... This control would be passed on to developers, giving us some better expectations of the equipment in use.
Sure, this might lead to a little stifling; I'm sure in each motherboard design session someone comes up with a reason to use Intel controllers or Via or nVidia, and in each case they're right, or not. Generally this doesn't affect the application developer; we tend to work far above the chipset controlling the PCI bus or soundcard. Sometimes, though, there are oddities that trickle up because of hardware differences beneath the OS for which we write.
If Apple does keep the mainboard and other hardware close to their chest, as they do now, the shift in CPU should make little difference to most of us. The Mac line will still only be sold at your local Apple retailer. The compilers and tools we use will probably look and act the same, just generating the code appropriate for the target processor.
In fact, it'll probably be the case, just like when they shifted from Motorola 68K chips to the PowerPC--software will be available for both for a while. I hope so; I'd hate to think my three-month old Mac Mini is already obsolete...
If you're looking for Mac on your Dell, go for Darwin, run X and KDE or Gnome, find a Carbon-like theme, and you'll be close. Yes, that works. If that's too weird,stick to a more comfortable BSD, LINUX, Solaris, or Windows environment on your Dell. If it's just not enough Mac for you, break down go get a Mac with a PowerPC chip.
Or wait until next year to see what they release.