Apple has already handled every transition in the book flawlessly why should this one be any different ?
Apple has.
But NeXT hasn't!
One cannot win the Windows/Intel war. Not on the same ground. Steve Jobs couldn't then. And he will not in the future.
It's more likely that developers will switch to Windows not users to Mac OS X.
And without Mac specific software from any vendor but Apple, Mac OS X will become insignificant. Like BeOS. Or others who also tried to compete with a better OS on Intel boxes.
Presuming that 'legacy' OS X software will run lacklustre in emulation on Intel hardware, users are forced into upgrading all their software.
At that point I might just 'upgrade' to Linux. At least that OS is free. And has lots more OpenSource software too.
The reasons for sticking with OS X would be very, very small.
People have been telling Apple to get out of the hardware business for a LONG TIME.
So what if Apple has been looking at the future Play Station and Xbox and it looks like they will be able run Mac software without much modification ?
Therefore, the only way to stay in the hardware business would be to SWITCH because the game boxes are TOO CHEAP.
Maybe they have some amazing new consumer toy and want to get out of the computer business, but before they quit they plan to produce some cool designer boxes and bring Mac OS X to Intel and then give it away IN THE END.
It will be a poison pill for MS OS sale !?!?!?!?!~~~
I put a suggestion in Apple's suggestion box a few days ago telling them to start using Intel processor, DO BOTH !
I have actually left the suggestion before but that was YEARS AGO.
Well, I guess they took my advice, if this ticks you off, SORRY !!!
I HAVE OFTEN SAID THEY ALWAYS TAKE MY ADVICE.
Why do people always worry about these transition ???
Apple has already handled every transition in the book flawlessly why should this one be any different ?
Trying to be funny doesn't help...
Apple has only ever underwent one majour transition, the 68k --> PPC one. The reason, and only reason for why that went as smoothly as it did was due to 1) the PPC being massevly more powerfull then the old 68k chips. Thus emulation of the old apps was easy, nowadays there isn't such a preformace gap to speak of. And I doubt that intel would be nice enough to implement special PPC ISA "swichover" instruktions like the PPC of yore did. 2) 68k development went on for a looong time. Again tis was possible due to the speed of which the 68k code ran on the PPC chips. A swichover today would have to be far far more "brutal", as doing PPC code would be vastly uneconomic, leaving people who aren't ready to "hustle" and swich swiftly to x86 by the wayside.
Then there will most likely be a vast, vast upswing of Mac On Linux users. All you'd need to run OSX, at full speed, would be a cheap and nasty x86 Linux box running MOL and a pirated copy of OSX.
But doom and gloom aside, does anybody know how much endian/ISA dependent code Apple has got in OSX. Disregarding the kernel stuff of course?
i'm really stunned. it will be nice to find out more details in the next few days. this is big news even for non-tech interests. there must be some kind of special Apple specific design that Intel will manufacture for this to work. what does this mean for Altivec? is there a truly serious PPC roadblock?
the transition is going to be fast. i wonder if there isn't going to be another refresh of the G5/Power Mac until then? this is odd. it seems interesting the timing is coinciding with the launch of Longhorn. could Apple be seeing the new look of the next PC evolution to be a greater threat than we thought. enough that Apple has to concede to the major architecture in order to compete...?
Apple has only ever underwent one majour transition, the 68k --> PPC one. The reason, and only reason for why that went as smoothly as it did was due to 1) the PPC being massevly more powerfull then the old 68k chips. Thus emulation of the old apps was easy, nowadays there isn't such a preformace gap to speak of. And I doubt that intel would be nice enough to implement special PPC ISA "swichover" instruktions like the PPC of yore did. 2) 68k development went on for a looong time. Again tis was possible due to the speed of which the 68k code ran on the PPC chips. A swichover today would have to be far far more "brutal", as doing PPC code would be vastly uneconomic, leaving people who aren't ready to "hustle" and swich swiftly to x86 by the wayside.
Then there will most likely be a vast, vast upswing of Mac On Linux users. All you'd need to run OSX, at full speed, would be a cheap and nasty x86 Linux box running MOL and a pirated copy of OSX.
But doom and gloom aside, does anybody know how much endian/ISA dependent code Apple has got in OSX. Disregarding the kernel stuff of course?
You are obviously quite young.
I tested many Apple prototypes, 040s , at Silicon Beach when they transitioned from 68000 to 68040.
But there was an OS transition before that they went from 16 bit to 32 bit, which I tested prototype Apple systems software and software compatiblity for.
Oh yah, I also handled sending Apple the "Compatibility Checker" CC information for Silicon Beach for the CC app which informed people as to what software was 32bit clean.
Then System 6 to OS 7 from basically beefed up Switcher tech, to multi-threaded multi-tasking.
Then OS 9 to Unix based ( Threaded Multitasking that people trust because it comes from Unix. ) OS X.
Oh and by the way, I wasn't trying, I was.
AND THEY DO ALWAYS DO WHAT I SUGGEST !!!
Oh, and I am also sure there were major transition before I understood what was going on, like the 128K Mac to 512 Mac Plus.
"Jobs? options are many. For one thing, Intel badly wants to sell its chips to Apple. For another, PC makers realize that the Mac OS X operating system is superior to Microsoft's Windows, and they want a piece of that market. FORTUNE has learned that Apple, Intel, and several PC companies already have the Mac OS X operating system working on Intel chips in their labs. And then there's the fact that more PC users are considering switching away from Windows. There are more Mac fanatics now than ever before. The Mac OS X operating system is superb, especially in its new ?Tiger? version, and Apple?s brilliant iPod is this decade?s signature tech device so far."
"Unlike IBM, Intel builds low-power, low-heat chips, especially for portable computers. (As I sit on a hotel bed in San Diego writing this column with my PowerBook G4 on my lap, I can definitely feel the heat.) This is critically important in an era when more and more PC buyers want laptops."
Then there will most likely be a vast, vast upswing of Mac On Linux users. All you'd need to run OSX, at full speed, would be a cheap and nasty x86 Linux box running MOL and a pirated copy of OSX.
But doom and gloom aside, does anybody know how much endian/ISA dependent code Apple has got in OSX. Disregarding the kernel stuff of course?
OS X on generic x86 hardware would need assloads of drivers. You're not going to get the hardware companies to write them, and it would require a lot of developers to be hired at apple. when a new user came across something that didn't work, he'd just say, "this apple stuff sucks. no hardware compatibility. i'm going back to windows".
I think OS X is completely endian-neutral by now (aside from OS 9/Classic code obviously). I think the Carbon API is platform netural; Cocoa obviously is. The entire OS, minus Classic, can probably run on x86 right now. I'm sure Apple has development builds in their labs. One of the ancient builds of OS X, Rhapsody DP2, ran on x86 just fine.
How is PIXAR going to subject you to their "MIND CONTROL MOVIE" and keep YOU under its spell if Apple has not already shipped Mac OS X for Intel Weenies !?!?!?!
Any Intel box where Mac OS X runs on would surely run Windows too. So that box could be booted in either Mac OS X or Windows.
Faced with that fact, why would a company like Microsoft (Office) or Alias (Maya) continue to develop their software for Mac OS X? Sure, it would be a much better user experience, and even a nicer development environment. But heck, it's extra effort too? Just tell the user to dual boot into Windows and use the Windows version instead.
The end of Mac OS X software...
And no matter what kind of tricks Apple will try, the moment OS X runs on Intel CPUs that can also run Windows, people will find a way to run OS X on any PC hardware.
Hmmm, bad news.
Apple can easily use an x86 processor with custom chips to support it - and nothing will make Windows run on that without special drivers. If Apple updates these custom chips periodically, even people who figure out the drivers for these custom chips would only be able to provide Windows running on old hardware. Hardly a benefit to run Windows like that.
OS X on generic x86 hardware would need assloads of drivers. You're not going to get the hardware companies to write them, and it would require a lot of developers to be hired at apple. when a new user came across something that didn't work, he'd just say, "this apple stuff sucks. no hardware compatibility. i'm going back to windows".
I think OS X is completely endian-neutral by now (aside from OS 9/Classic code obviously). I think the Carbon API is platform netural; Cocoa obviously is. The entire OS, minus Classic, can probably run on x86 right now. I'm sure Apple has development builds in their labs. One of the ancient builds of OS X, Rhapsody DP2, ran on x86 just fine.
You do bring up a very good point about needing a ton of drivers out of the gate to keep a userbase. however whats the chances of being able to use FreeBSD driver base ? and or port over linux and other drivers out there?
Interesting article. We may see some light at the WWDC. Apple should stay with IBM, they developed dual core first why not stick with em? PPC chips have proven to better with graphic intense software.
If on Monday Apple announces that it is "scrapping" its partnership with IBM and switching to Intel, then Apple must be prepared to swallow a large short-term decline in hardware sales. Why stick with the old horse when a new one is in the offing? Clued-in consumers may question the general stability of the platform itself, particularly if stories leak later on about Steve's (real or imaged) ire towards IBM as the "irrational" impetus for such a switch. From a PR standpoint, going with Intel seems to have many more cons than pros. I hope the folks at Cupertino have hashed this one out from every angle.
You do bring up a very good point about needing a ton of drivers out of the gate to keep a userbase. however whats the chances of being able to use FreeBSD driver base ? and or port over linux and other drivers out there?
The FreeBSD stuff in OS X is only user-level and API level. The kernel is called XNU and is based on the Mach kernel, developed at Carnegie Mellon, Avie Tevanian's alma matter. However, OS X uses a unique driver system called the I/O Kit. I/O Kit drivers are object-oriented and written in a subset of C++. Very different from anything out there. Porting Linux drivers is a possibility, but what about the GPL issues? Maybe FreeBSD drivers would work since they're under the BSD license. It would certainly involve some work...
Iswallow a large short-term decline in hardware sales. Why stick with the old horse when a new one is in the offing? Clued-in consumers may question the general stability of the platform itself, particularly if stories leak later on about Steve's (real or imaged) ire towards IBM as the "irrational" impetus for such a switch. From a PR standpoint, going with Intel seems to have many more cons than pros.
That kind of thing has happened in the past. It has almost killed SGI (they're basically dead anyway); it's killing HP's Unix server market. You know how the press likes to go after Apple. Apple would get hammered and hardware sales would plummet, if history is any indication. I agree, the cons of an Intel switch likely outweigh the pros. You know the old adage, those who ignore history are doomed to repeat it. Of course we are dealing with Jobs' ego here.
Comments
Apple has already handled every transition in the book flawlessly why should this one be any different ?
Apple has.
But NeXT hasn't!
One cannot win the Windows/Intel war. Not on the same ground. Steve Jobs couldn't then. And he will not in the future.
It's more likely that developers will switch to Windows not users to Mac OS X.
And without Mac specific software from any vendor but Apple, Mac OS X will become insignificant. Like BeOS. Or others who also tried to compete with a better OS on Intel boxes.
Presuming that 'legacy' OS X software will run lacklustre in emulation on Intel hardware, users are forced into upgrading all their software.
At that point I might just 'upgrade' to Linux. At least that OS is free. And has lots more OpenSource software too.
The reasons for sticking with OS X would be very, very small.
So what if Apple has been looking at the future Play Station and Xbox and it looks like they will be able run Mac software without much modification ?
Therefore, the only way to stay in the hardware business would be to SWITCH because the game boxes are TOO CHEAP.
Maybe they have some amazing new consumer toy and want to get out of the computer business, but before they quit they plan to produce some cool designer boxes and bring Mac OS X to Intel and then give it away IN THE END.
It will be a poison pill for MS OS sale !?!?!?!?!~~~
Originally posted by MACchine
I put a suggestion in Apple's suggestion box a few days ago telling them to start using Intel processor, DO BOTH !
I have actually left the suggestion before but that was YEARS AGO.
Well, I guess they took my advice, if this ticks you off, SORRY !!!
I HAVE OFTEN SAID THEY ALWAYS TAKE MY ADVICE.
Why do people always worry about these transition ???
Apple has already handled every transition in the book flawlessly why should this one be any different ?
Trying to be funny doesn't help...
Apple has only ever underwent one majour transition, the 68k --> PPC one. The reason, and only reason for why that went as smoothly as it did was due to 1) the PPC being massevly more powerfull then the old 68k chips. Thus emulation of the old apps was easy, nowadays there isn't such a preformace gap to speak of. And I doubt that intel would be nice enough to implement special PPC ISA "swichover" instruktions like the PPC of yore did. 2) 68k development went on for a looong time. Again tis was possible due to the speed of which the 68k code ran on the PPC chips. A swichover today would have to be far far more "brutal", as doing PPC code would be vastly uneconomic, leaving people who aren't ready to "hustle" and swich swiftly to x86 by the wayside.
Then there will most likely be a vast, vast upswing of Mac On Linux users. All you'd need to run OSX, at full speed, would be a cheap and nasty x86 Linux box running MOL and a pirated copy of OSX.
But doom and gloom aside, does anybody know how much endian/ISA dependent code Apple has got in OSX. Disregarding the kernel stuff of course?
the transition is going to be fast. i wonder if there isn't going to be another refresh of the G5/Power Mac until then? this is odd. it seems interesting the timing is coinciding with the launch of Longhorn. could Apple be seeing the new look of the next PC evolution to be a greater threat than we thought. enough that Apple has to concede to the major architecture in order to compete...?
there are a lot of questions
Originally posted by Eric_Z
Trying to be funny doesn't help...
Apple has only ever underwent one majour transition, the 68k --> PPC one. The reason, and only reason for why that went as smoothly as it did was due to 1) the PPC being massevly more powerfull then the old 68k chips. Thus emulation of the old apps was easy, nowadays there isn't such a preformace gap to speak of. And I doubt that intel would be nice enough to implement special PPC ISA "swichover" instruktions like the PPC of yore did. 2) 68k development went on for a looong time. Again tis was possible due to the speed of which the 68k code ran on the PPC chips. A swichover today would have to be far far more "brutal", as doing PPC code would be vastly uneconomic, leaving people who aren't ready to "hustle" and swich swiftly to x86 by the wayside.
Then there will most likely be a vast, vast upswing of Mac On Linux users. All you'd need to run OSX, at full speed, would be a cheap and nasty x86 Linux box running MOL and a pirated copy of OSX.
But doom and gloom aside, does anybody know how much endian/ISA dependent code Apple has got in OSX. Disregarding the kernel stuff of course?
You are obviously quite young.
I tested many Apple prototypes, 040s , at Silicon Beach when they transitioned from 68000 to 68040.
But there was an OS transition before that they went from 16 bit to 32 bit, which I tested prototype Apple systems software and software compatiblity for.
Oh yah, I also handled sending Apple the "Compatibility Checker" CC information for Silicon Beach for the CC app which informed people as to what software was 32bit clean.
Then System 6 to OS 7 from basically beefed up Switcher tech, to multi-threaded multi-tasking.
Then OS 9 to Unix based ( Threaded Multitasking that people trust because it comes from Unix. ) OS X.
Oh and by the way, I wasn't trying, I was.
AND THEY DO ALWAYS DO WHAT I SUGGEST !!!
Oh, and I am also sure there were major transition before I understood what was going on, like the 128K Mac to 512 Mac Plus.
"Jobs? options are many. For one thing, Intel badly wants to sell its chips to Apple. For another, PC makers realize that the Mac OS X operating system is superior to Microsoft's Windows, and they want a piece of that market. FORTUNE has learned that Apple, Intel, and several PC companies already have the Mac OS X operating system working on Intel chips in their labs. And then there's the fact that more PC users are considering switching away from Windows. There are more Mac fanatics now than ever before. The Mac OS X operating system is superb, especially in its new ?Tiger? version, and Apple?s brilliant iPod is this decade?s signature tech device so far."
"Unlike IBM, Intel builds low-power, low-heat chips, especially for portable computers. (As I sit on a hotel bed in San Diego writing this column with my PowerBook G4 on my lap, I can definitely feel the heat.) This is critically important in an era when more and more PC buyers want laptops."
Originally posted by Eric_Z
Then there will most likely be a vast, vast upswing of Mac On Linux users. All you'd need to run OSX, at full speed, would be a cheap and nasty x86 Linux box running MOL and a pirated copy of OSX.
But doom and gloom aside, does anybody know how much endian/ISA dependent code Apple has got in OSX. Disregarding the kernel stuff of course?
OS X on generic x86 hardware would need assloads of drivers. You're not going to get the hardware companies to write them, and it would require a lot of developers to be hired at apple. when a new user came across something that didn't work, he'd just say, "this apple stuff sucks. no hardware compatibility. i'm going back to windows".
I think OS X is completely endian-neutral by now (aside from OS 9/Classic code obviously). I think the Carbon API is platform netural; Cocoa obviously is. The entire OS, minus Classic, can probably run on x86 right now. I'm sure Apple has development builds in their labs. One of the ancient builds of OS X, Rhapsody DP2, ran on x86 just fine.
Originally posted by hobBIT
My main concern would be software availability.
Any Intel box where Mac OS X runs on would surely run Windows too. So that box could be booted in either Mac OS X or Windows.
Faced with that fact, why would a company like Microsoft (Office) or Alias (Maya) continue to develop their software for Mac OS X? Sure, it would be a much better user experience, and even a nicer development environment. But heck, it's extra effort too? Just tell the user to dual boot into Windows and use the Windows version instead.
The end of Mac OS X software...
And no matter what kind of tricks Apple will try, the moment OS X runs on Intel CPUs that can also run Windows, people will find a way to run OS X on any PC hardware.
Hmmm, bad news.
Apple can easily use an x86 processor with custom chips to support it - and nothing will make Windows run on that without special drivers. If Apple updates these custom chips periodically, even people who figure out the drivers for these custom chips would only be able to provide Windows running on old hardware. Hardly a benefit to run Windows like that.
Originally posted by unixguru
OS X on generic x86 hardware would need assloads of drivers. You're not going to get the hardware companies to write them, and it would require a lot of developers to be hired at apple. when a new user came across something that didn't work, he'd just say, "this apple stuff sucks. no hardware compatibility. i'm going back to windows".
I think OS X is completely endian-neutral by now (aside from OS 9/Classic code obviously). I think the Carbon API is platform netural; Cocoa obviously is. The entire OS, minus Classic, can probably run on x86 right now. I'm sure Apple has development builds in their labs. One of the ancient builds of OS X, Rhapsody DP2, ran on x86 just fine.
You do bring up a very good point about needing a ton of drivers out of the gate to keep a userbase. however whats the chances of being able to use FreeBSD driver base ? and or port over linux and other drivers out there?
Originally posted by DHagan4755
http://www.eweek.com/article2/0,1759,1823273,00.asp
Interesting article. We may see some light at the WWDC. Apple should stay with IBM, they developed dual core first why not stick with em? PPC chips have proven to better with graphic intense software.
Originally posted by oh_the_humanity
You do bring up a very good point about needing a ton of drivers out of the gate to keep a userbase. however whats the chances of being able to use FreeBSD driver base ? and or port over linux and other drivers out there?
The FreeBSD stuff in OS X is only user-level and API level. The kernel is called XNU and is based on the Mach kernel, developed at Carnegie Mellon, Avie Tevanian's alma matter. However, OS X uses a unique driver system called the I/O Kit. I/O Kit drivers are object-oriented and written in a subset of C++. Very different from anything out there. Porting Linux drivers is a possibility, but what about the GPL issues? Maybe FreeBSD drivers would work since they're under the BSD license. It would certainly involve some work...
Oh wait....
Originally posted by Sybaritic
Iswallow a large short-term decline in hardware sales. Why stick with the old horse when a new one is in the offing? Clued-in consumers may question the general stability of the platform itself, particularly if stories leak later on about Steve's (real or imaged) ire towards IBM as the "irrational" impetus for such a switch. From a PR standpoint, going with Intel seems to have many more cons than pros.
That kind of thing has happened in the past. It has almost killed SGI (they're basically dead anyway); it's killing HP's Unix server market. You know how the press likes to go after Apple. Apple would get hammered and hardware sales would plummet, if history is any indication. I agree, the cons of an Intel switch likely outweigh the pros. You know the old adage, those who ignore history are doomed to repeat it. Of course we are dealing with Jobs' ego here.
originally posted by chris v:
A relevant question would be-- could Apple survive two years without selling a single powermac?
Good companies exist to thrive, not merely survive, but your question is an essential one if Jobs really plans to make the jump.