Mac Musings

Left Behind by Mac OS X or Up to Date with Linux?

Daniel Knight - 2008.06.04 -

I have very mixed feelings about Mac OS X 10.6, whatever kind of cat they name it after.

Rumor has it that it will be an Intel-only operating system, leaving all PowerPC Macs behind. And it will be demonstrated and an early version seeded at the Worldwide Developer Conference (WWDC). And it should be out before the 2010 WWDC.

Cutting Off PowerPC

It's inevitable that Apple is someday going to drop support for PowerPC Macs in its operating system. Apple added Intel support to Tiger (Mac OS X 10.4) in 2006, and Leopard (10.5) is the first fully universal version of the Mac OS, supporting both Intel and PowerPC processors.

Leopard, released in November 2007, showed us how far Apple was willing to go in cutting off support for recent models. The oldest Mac officially supported was the July 2001 867 MHz Quicksilver Power Mac G4; the most recently discontinued unsupported model was the October 2003 800 MHz iBook G4, replaced by a 1 GHz model in April 2004. While Leopard can run on Macs going back to 1999 - using a firmware hack and as long as they have a G4 CPU and AGP graphics - officially supported Macs ranged from 43 months old (end-of-life 800 MHz iBooks) to 76 months old (the 867 MHz Quicksilver Power Mac).

The first Intel-based Macs were introduced in January 2006, and it wasn't until August 2006 that the Mac Pro was introduced to replace the last generation Power Mac G5. I honestly can't see Apple discontinuing support for a 3-year-old computer - 3-1/2 years minimum, and 4 years is more likely. That would put the release of the first Intel-only version of Mac OS X in the January 2010 Macworld Expo to June 2010 WWDC range.

A very late 2009 release (October or November) is the earliest I can imagine Apple abandoning PowerPC with a new version of OS X - maybe two years after the introduction of Leopard.

Mac OS X 10.6

Apple got a lot of mileage out of Tiger, which was on the market for 2-1/2 years. It supported G3 Macs back to the Cro-Magnon era (er, the January 1999 blue & white Power Mac G3), and it was the first version of OS X compiled and sold for Intel Macs. Tiger was relatively bug-free at release, and 10.4.1 nailed most of the problems. Tiger set the bar for quality and stability.

In six months, Leopard has already had three significant upgrades, and it's finally beginning to approach the stability of 10.4.0 and 10.4.1. Mac users are already making lists of things that should be fixed in 10.5.4. Some feel like beta testers, and some are reverting to Tiger because of stability issues with Leopard.

Best of all (for Tiger users), Apple has continued to provide updates to the last major version of OS X at least going back to the 10.3 era. That meant security updates and minor fixes for 10.2.8 back then - and it means the same support for 10.4.11 until 10.6 is released (and possibly for a brief period after that).

It's nice that Apple has a road map for 10.6, and I think it would be great if they previewed it at this summer's WWDC. But I hope it won't turn into Apple's "Windows 7", the code name for the first post-Vista release of Windows. Microsoft won't admit that Vista is a bust, but they've created yet another loophole for continued sales of Windows XP while doing proof of concept demonstrations of some Windows 7 features.

Leopard isn't the bust that Vista is, but despite 2-1/2 years between the release of 10.4 and 10.5, the sheer number of bugs and problems makes it appear that Apple rushed to market. Mac users are used to higher standards; let's hope that Apple learned its lesson with the extent of Leopard updates and gets a lot more bugs out of 10.6 before releasing it.

If 10.6 ships in late 2009, it's not unthinkable that it would be an Intel-only release. However, if it ships in mid-2009, that's awfully early to be pulling the plug on PowerPC support. I'm betting that if 10.6 is just for Intel Macs, it won't be out until late 2009.

The Lion's Share

We haven't yet heard a code name for OS X 10.6, so we'll propose Lion. Macs already have the lion's share of the non-Windows operating system market, and an Intel-only version that could be easily hacked to boot on standard PC hardware could help Apple cut into both the Linux and Windows markets.

Speaking of Linux, there's a whole different way of building an operating system: Instead of tying the interface and operating system together, Linux has an underlying operating system that runs on just about anything. There's no 867 MHz G4 cutoff, as with Leopard, or 1 GHz, as with Vista. Nor do you need 20 GB of hard drive space or 512 MB of RAM. Linux can run on a 386 CPU, fit in as little as 100 MB of drive space, and function with less than 16 MB of RAM (some gurus claim as little as 4 MB). You can run Linux from a flash drive or CD.

It's up to the user, not the creator of the operating system, to determine what graphical user interface (if any) is used. With a GUI, Linux becomes more demanding - a Pentium III or later with 256 MB of RAM and 1 GB of available hard drive space is more than sufficient for typical tasks such as word processing, email, spreadsheets, and browsing the Web. But those are just guidelines, not requirements etched in stone or built into an installer program.

In the end, you get to decide how sophisticated a user interface you want to use and how much power you want to use. This is totally contrary to the way Microsoft and Apple work, always demanding more resources and creating operating systems that simply overwhelm old hardware.

An Up-to-Date OS

I wish Apple had as enlightened an approach to Mac OS X as the Linux community has toward its operating system. What if you could run core OS X on any PowerPC Mac and then choose how sophisticated a user interface you wanted to use? If Apple had elected to decouple the UI from the OS, ancient Macs could run OS X from the command line, newer ones with a less sophisticated GUI (along the lines of the Classic Mac OS), G3 models something akin to Tiger, and G4 and later Macs the full Leopard experience.

For some reason, Apple chose not to do that - and from a marketing standpoint, it probably makes more sense to have a single unified OS/UI. But from a low-end perspective, the Linux approach makes more sense - one core operating system that can run on almost any hardware and a range of user interfaces we can choose between.

Yeah, it makes things more complex, but it eliminates that ongoing headaches and debates as Apple phases out support for one generation of hardware after another. Instead of not supporting G3s with Leopard, the underlying OS could have run on G3 Macs while the Leopard interface required a G4.

Why is that important? Because the Apple and Microsoft approach leaves perfectly good hardware and its users behind. We've seen it again and again: Classic Mac OS users without a truly modern browser, early versions of OS X no longer being supported in Firefox, Safari abandoning Mac OS X 10.4.10 and earlier with version 3.0, etc.

Why should you have to have the latest version of the operating system to have a fully up-to-date browser? If Camino can provide an up-to-date experience with Mac OS X 10.3.9, why can't Safari? Why does Firefox 2 only require Mac OS X 10.2.x, while Firefox 3 needs 10.4 or later?

Yes, I realize that software development is easier with newer tools, some of which don't support older versions of operating systems. But I think the Linux world is on to something: free upgrades and minimal hardware requirements.

While some Mac users will gladly move forward to OS X 10.6 when it ships, just as many have already embraced Leopard, some will be left behind - and as their version of Mac OS X 10.2, 10.3, 10.4, or 10.5 grows increasingly outdated by new apps, we may see more and more low-end Mac users looking into Linux (which is free and offers up-to-date apps) as a viable alternative to Apple's operating systems on their aging hardware.

Funny, isn't it, how Apple will leave your hardware behind while the Linux community will keep you up to date.