12 Jun
2009
12 Jun
'09
8:13 a.m.
At 1:50 PM +0100 6/12/09, Liam Proven wrote:
2009/6/11 Zane H. Healy <healyzh at
aracnet.com>:
You might have a point, with Prelude to Rhapsody I built a Pentium
box to run it, I can't remember if the Pentium II was out yet or not.
The Pentium Pro was capable of running a pretty nice system, but it
was more than I wanted to spend. I quickly upgraded. When they
bought NeXT, the Pentium 4 was underdevelopment, and it was out
before Mac OS X, IIRC. The two architectures leapfrogged each other
a time or two before x86 left them in the dust, but lets not forget
the years of bad performance with the G4. Realistically the PPC
based Mac's spent more time trailing x86 boxes than ahead of them.
Still there is no denying the greatness of the G5, especially as I'm
typing this on a G5 dual 2Ghz. With the move to EM64T the distance
has been even greater.
The only thing Apple did wrong with the change
from PPC to Intel was that
they took to long. They wasted a lot of resources porting the OS from
68k/x86/SPARC/PA-RISC to PPC. They should have focused their efforts on
getting it running on x86. Remember the first developer releases of
Rhapsody ran on x86. Those responsible for PPC simply were not responsive
enough to Apple's or its customers needs.
Zane
That's just not realistic. They needed a new OS, not only for their
new hardware, but for the millions of existing PowerPC machines they
had out there. What's more, they also had thousands of existing
PowerPC applications that they needed to run, too.
In 1996 when Apple acquired NeXT, the lackluster Pentium Pro was the
state of the x86 art. A migration then would have been disastrous -
the PPro was not even that competitive with the best PowerPCs,
certainly could not have emulated the PPC effectively. PowerPC
remained highly competitive with the Pentium II, PIII and trounced the
wretched Pentium 4. And OpenStep was portable - it already supported MC68K
and x86-32 in
shipping versions, and as you say, SPARC in the labs. I don't recall a
PA-RISC version, but I'm sure you're right.
All 4 versions shipped as part of OPENSTEP 4.2, and I believe date
back to 4.0 (not possitive).
They couldn't have done it any earlier. Indeed, I
think it would
almost have been better to have gone straight for x86-64 throughout
the range, and for Apple to have skipped the handful of early Core
1-based models, which are 32-bit only.
But Apple *had* to wait until the bulk of its userbase and developers
and apps were on OS X, happy to move to a new version - the first
version to officially support x86 was 10.4.5 at the start of 2006. The
other enabling technology was Transitive's QuickTransit PowerPC
emulator, so that the millions of users could still use PowerPC OS X
apps on their new Macintels. QuickTransit wasn't around 'til 2004 at
the earliest.
In summary, no, I entirely disagree. Apple jumped at just about the
right time, and could not have done it much earlier. Before the Core
processors, the performance advantage just wasn't there, and before
2006 or so, too many people were still running Classic applications,
which wouldn't work on OS X/86.
I'll agree with some of these arguments, but I still think that the
change should have occurred in the G4 days at the latest. They did
pick a great time to jump though with the release of the Core CPU's.
Zane
--
| Zane H. Healy | UNIX Systems Administrator |
| healyzh at aracnet.com (primary) | OpenVMS Enthusiast |
| MONK::HEALYZH (DECnet) | Classic Computer Collector |
+----------------------------------+----------------------------+
| Empire of the Petal Throne and Traveller Role Playing, |
| PDP-10 Emulation and Zane's Computer Museum. |
| http://www.aracnet.com/~healyzh/ |