<<< Date Index >>>     <<< Thread Index >>>

more on Opinions Re: [IP] Apple to ditch IBM, switch to Intel chips rumor on CNet



Begin forwarded message:

From: Roman Gollent <roman-ip@xxxxxxxxxxx>
Date: June 4, 2005 3:48:03 PM EDT
To: David Farber <dave@xxxxxxxxxx>
Subject: Re: more on Opinions Re: [IP] Apple to ditch IBM, switch to Intel chips rumor on CNet 



Two words: Fat Binaries

Those of us that use(d) NeXTStep/OpenStep on both "black" (68k) and
"white" (x86) machines have already been around for one such
architectural transition, and this was long before the days of VMWare et
al.  The idea is that every application binary is compiled to run on
both architectures.  This allowed us to compile every application
running in the organization once (per code rev) and share binaries via
NFS to either type of box without adding any discriminating intelligence
on the client side.  It's not a far stretch of the imagination that
Apple would be able to redeploy this technology on MacOS.

No emulation penalties and few, if any, complications on the user side.

Roman


On Sat, Jun 04, 2005 at 02:11:39PM -0400, David Farber wrote:




Begin forwarded message:

From: Matt Clauson <mec@xxxxxxxxxx>
Date: June 4, 2005 11:32:29 AM EDT
To: dave@xxxxxxxxxx
Subject: Opinions Re: [IP] Apple to ditch IBM, switch to Intel chips
rumor on CNet


On Jun 4, 2005, at 6:39 AM, David Farber wrote:




http://news.com.com/2102-1006_3-5731398.html?tag=st.util.print




I have significant doubts that Apple as a corporation is stupid
enough to shoot themselves in the foot like this.  Anyone who follows
the more hardcore tech rags, or even a news aggregator like Slashdot,
will remember the recent PearPC/CherryOS debacle, where the specifics
of emulation of grossly different hardware (in this case, the PowerPC
CPU architecture on the Intel x86 platform) can have massively
horrible results in performance.  In this case, it's allegedly due to
the horribly different CPU designs, including instruction sets and
number of registers between the two architectures.  Apparently the
reverse type of emulation (x86 emulation on PPC architecture) doesn't
have as many problems.

The major question to ask is what will Apple's user base do if this
switch happens?  A lot of current (and theoretically non-legacy)
applications will have to run under the emulation, significantly
impairing performance.  Does Apple expect that the users will shell
out hundreds (if not thousands) of dollars to upgrade their
applications to something that will run natively on the new
platform?  What about the thousands of dollars in man-hours for third-
party software vendors (like Adobe) to port their products to OSX/
Intel platforms?  While users pay a premium for Apple hardware
currently (and I'll admit, I've been bitten by the bug recently and
am beginning my own conversion after over a decade of almost-rabid
Linux usage) that premium is usually well justified in the elegance
and performance of the hardware, especially the past few years under
OSX.  Does Apple really expect their users to end up paying thousands
of dollars to switch to what seems to me like a computationally
inferior platform?

I pray that they don't, for this would almost be a form of corporate
seppuku if they do.

--mec


-------------------------------------
You are subscribed as roman@xxxxxxxxx
To manage your subscription, go to
http://v2.listbox.com/member/?listname=ip
Archives at: http://www.interesting-people.org/archives/interesting- people/ 


-------------------------------------
You are subscribed as roessler@xxxxxxxxxxxxxxxxxx
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/