Attention owners of the original Intel Macs: If you have not already
done so, you might want to upgrade your systems to Mac OS X 10.6 "Snow Leopard".
It may well be the last Mac OS that you will be able to run on your Core
Solo or Core Duo system.
Each version of Mac OS X, with the exception of Puma (10.1) and
Jaguar (10.2), has increased
the hardware requirements for installation beyond those of its
predecessor, essentially obsoleting another group of Macs from running
the current version of OS X (more on this below).
The following chart is a timeline of each major version of Mac
OS X up to the present, along with its release date and a brief
description of the hardware that particular version "left behind". The
last column lists the minimum potential age of the obsoleted hardware.
In other words, this column reflects the latest date you could have
bought a new (current model) Mac and have it be officially unsupported
by the corresponding version of Mac OS X. In some cases, there may
have been more than one obsoleted Mac available on that date, but for
simplicity, I only listed one representative model. For example, the
last date you could have purchased a PowerBook G3 "Kanga" as new was May
4, 1998. Had you done so, when Apple released Mac OS X 10.0
"Cheetah" 2 years, 10 months, and 21 days later, you would have found
your PowerBook unsupported by the then-current version of the Mac
OS.
A Brief History of Mac OS X
| Version |
Code Name |
Release Date |
Hardware Left
Behind |
Time Until
Obsolescence |
| 10.0 |
Cheetah |
Mar 24, 2001 |
Pre-G3 PPC |
2 years, 10 months, 21
days (PowerBook G3 Kanga; May 4, 1998) |
| 10.1 |
Puma |
Sep 25, 2001 |
Same as 10.0 |
3 years, 4 months, 22
days (PowerBook G3 Kanga; May 4, 1998) |
| 10.2 |
Jaguar |
Aug 24, 2002 |
Same as 10.0 |
4 years, 3
months, 21 days (PowerBook G3 Kanga; May 4, 1998) |
| 10.3 |
Panther |
Oct 24, 2003 |
G3 systems without
built-in USB* (Old-World ROM G3) |
4 years, 9 months, 24
days (Beige G3s; Jan 1,
1999) |
| 10.4 |
Tiger |
Apr 29, 2005 |
New-World ROM G3
without built-in FireWire* |
4 years, 7 months, 17
days (clamshell
iBook; Sep 13, 2000) |
| 10.5 |
Leopard |
Oct 26, 2007 |
all G3; G4 below 867
MHz |
3 years, 6 months, 8 days
(12" iBook G4/800; Apr
19, 2004) |
| 10.6 |
Snow Leopard |
Aug 28, 2009 |
All PowerPC |
3 years, 22 days
(Power Mac G5 Quad; Aug
7, 2006) |
| 10.7 |
Cougar?
Lynx?
Lion? |
unknown |
Intel CoreDuo / Core Solo
systems? |
As of Jan 19, 2010: 2
years, 5 months, 13 days (Mac
Mini Core Duo; Aug 7, 2007) |
Unsupported Doesn't Mean Impossible
Much of the hardware left behind appears to be based on arbitrary
decisions by Apple rather than technical limitations. For example, many
pre-G3 PowerPC systems can be made to run Mac OS X versions up
through and including Jaguar, and all G3-based Macs (except the
PowerBook G3 "Kanga") can run OS X versions up to and including
Tiger through the use of third-party installers such as XPostFacto.
Sub-867 MHz G4 Macs can run Leopard by entering a command in the Mac's
Open Firmware interface to fool the Leopard Installer into thinking
it has a clock rate of 867 MHz or greater.
Except for features requiring specific hardware (e.g., graphics
acceleration, DVD burning), the operating system offers the same
functionality as on the officially supported hardware. Leopard is a
partial exception to this rule. It cannot be made to run stably on G3
systems, as the G3 processor lacks AltiVec, a requirement for 10.5 on
PowerPCs. The 867 MHz minimum for a G4 system is, however,
artificial.
Intel Is Absolute
That's where the comparison ends. Nothing can be done to make
OS X 10.6 Snow Leopard run on any PowerPC hardware. It is a
technical impossibility, as the PowerPC and Intel processor families
are fundamentally different in their architecture. This is not an
arbitrary decision; it really is a technical limitation.
The chart above is somewhat deceiving. At first glance, it appears
that Apple's officially supported hardware remained unchanged through
three iterations of Mac OS X. However, anyone who has used
OS X from the beginning, as I have (I bought the public beta on
the day of its release), knows that this is misleading. Cheetah and
Puma were really just extensions of the public beta.
It was not until the release of Jaguar that most people considered
Mac OS X ready for prime time. It was only with the release of Jaguar
that many Mac users started booting OS X as the default OS.
Up until its
release, most of us continued to boot into OS 9 (though Apple made
OS X the default startup OS on new Macs beginning with the release
of the 10.1.2 update). As a side note, the Happy Mac, which had
appeared during the Mac OS startup sequence for almost 18 years, was
replaced with a large grey Apple logo in Mac OS X 10.2. Whether
intentional or not, this clearly signaled the beginning of a new
era.
Development of the Mac OS brought usability up to mainstream
standards in Jaguar. With the next major release, Panther, the time
span encompassing supported hardware peaked at a record of 4 years, 9
months, 24 days. From that point, there has been a steady downward
crawl in the officially supported life span of Macs.
Anticipating OS X 10.7
If Mac OS X 10.7 were released today (Jan 19, 2010) and if it
supported only 64-bit Intel processors (the Core 2 Duo models), the
minimum potential age of the obsoleted hardware would be 2 years, 5
months, 13 days. However, if 10.7 were released on Aug 28, 2010,
exactly one year after the release of 10.6, the minimum potential age
of the obsoleted hardware would be 3 years, 22 days - exactly the same
age as the hardware obsoleted by 10.6 and well within the curve of
where the phase-out time span is trending.
Obsoleting hardware with newer operating systems does not
necessarily make for an evil company, as long as it does so on a
reasonable schedule. The new Mac you buy today should not be
unsupported next month, but is it reasonable to expect full support
five, six, seven, or more years from now?
From time to time, Apple must leave users behind for both technical
and practical reasons. From a technical standpoint, continuing to
support legacy hardware far beyond its otherwise useful life span leads
to OS bloat, which leads to bugs, which leads to an unstable OS, and
pretty soon you have Microsoft Windows. This is what has happened with
Windows over the years, as Microsoft continued to build in support for
programs that originally ran under long-dead operating systems of the
1980s just to appease a few business customers.
There is also a practical reason for putting older hardware to bed.
Apple has a business model that is largely different from Microsoft.
Apple makes far more money on its hardware sales than it does from OS
sales. When an OS no longer supports certain older hardware, the owner
of that older hardware is much more likely to buy new hardware from
Apple, thus placing cash in its coffers.
Apple is a unique company, but it is not exempt from the need to
make money. It is this money that fuels Apple's research and
development, which leads to the cutting-edge innovations that make me
want to buy and use Apple's products. In short, this is called
survival. And that's a Good Thing.
The Way Ahead
Several factors point toward Apple leaving behind the original Intel
Macs. First, it is unlikely that Mac OS X 10.7 would be released a
mere 12 months after 10.6. Since 10.2, the shortest span between major
OS releases has been 14 months, and the last two releases have hovered
closer to the two year mark. The later the release date of 10.7, the
more likely it is that the 32-bit Intel Macs won't be invited to the
party.
Most importantly, there could easily be compelling technical reasons
to exclude the 32-bit processors. Snow Leopard is already fully 64-bit
capable. With the exception of the Xserve, all current Macs ship from the
factory configured to boot into 32-bit mode by default. The user can
easily change this, but 32-bit is the default. This is primarily for
compatibility with existing 32-bit hardware drivers and some finicky
32-bit software.
Most 32-bit programs run perfectly fine when Snow Leopard's 64-bit
kernel is used. By 10.7's probable release date, there will most likely
be very few programs or drivers still around that won't either be
64-bit native or, at the very least, well-behaved when run on a 64-bit
kernel. In short, the necessity of supporting a 32-bit processor will
be minimal.
Phasing out 32-bit processor support will also give the Mac OS
development team some well-deserved and much needed breathing room.
Once 32-bit support is gone, there are no more major technological
challenges on the horizon. 64-bit processors are not likely to give way
to another architecture in the near future.
Apple's engineers have probably been working overtime the last
several years supporting both 32- and 64-bit PowerPC G5 processors, as
well as 32- and 64-bit Intel processors. With only 64-bit Intel on the
table, the engineers can devote their time to other pursuits,
presumably many of which will enhance the end-user experience.
Counterpoint
There is one consideration weighing against casting out 32-bit Intel
in the next OS release. To drop support for any Intel processors
so soon after Apple moved to Intel could be seen as a slap in the face
by some of Apple's most valuable customers: the early adopters. Apple
might think twice about abandoning those who bought Intel Macs the
minute they were released. Had Intel Macs not moved well in the
beginning, the Intel transition could have hit a few potholes. Apple
needed to sell Macs to fund further OS development. Had Apple's user
base forced the cart before the horse, i.e., not bought Macs until
there was further OS developments, Apple could have had some financial
hard times. Fortunately, there were users ready, willing, and able to
purchase the first Intel Macs, and they came with cash in hand. Apple
cannot lightly brush aside these users.
Counter-counterpoint
Even this consideration has its counterpoint. By this time, many of
the original Intel Macs have moved on to subsequent users. Thus Apple
won't really risk alienating very many early adopters, as this group
tends to buy new Macs at a quick pace anyway and probably no longer own
their original Intel Macs.
We bought an Intel Core
Duo iMac and MacBook Pro, both
refurbished, from the Apple Store shortly after the introduction of
each model's successor. Since that time, both of our Macs have moved on
to new homes and have been replaced by a MacBook Air and newer MacBook
Pro. (My wife and I don't generally consider ourselves the
cutting-edge, early adopter type. This is borne out by the Power Mac G4 Cube I still use
daily, and the fact that you are reading this article on Low End
Mac.)
An Ironic Twist
The truly ironic situation would be users who purchased a 64-bit G5
iMac or Power Mac, then "upgraded" when the first Intel Macs, all
32-bit, were released, and finds themselves out in the cold when Mac
OS X 10.7 does not support their Macs because they do not have
64-bit processors.
There probably aren't many of these people now, but you can bet
there are at least a few.
Intel Macs left behind in the next version of Mac OS X? I'd bet
on it. 
