[osxisfun]

Originally Posted by sideus

If I was a developer after gone through the OS 9 to OS X transistion and having to do it again for an Intel based system, I'd be extremely pissed.

unless it was relatively easy (relatively).

say like transitive or xcode -> target PC OSX.

I'm not saying its going to be a picnic but don't you think apple has thought long and hard about this if its true?

------------

Either scoble is being taken for a ride or we are in for an exciting monday:

I'm not wrong. Wait until Monday. Apple is going to Intel x86 chipset for OSX and dumping PowerPC. The News.com report is very accurate.

There are major changes coming.

I don't report such things lightly.
Posted by Robert Scoble on June 04, 2005 at 12:37 PM EDT
Website: http://scoble.weblogs.com #

[zoetrope]

Yep, there's going to be a helluva lot of developer's p!$$ed. There's going to be even more end user's p!$$ed!

But there's no more powerful a force for creating the RDF than Hollywood. And Hollywood is the reason for all of this.

Anybody now regretting Steve Jobs ever bought Pixar from George Lucas?

[Boochie]

New York Times has an article dated Monday on the story:

http://nytimes.com/2005/06/06/technology/06apple.html

The intrigue builds. The NYTimes article doesn't mention the Transitive technology, though.

[osxisfun]

This part from the article adds to the mix as well:

crosoft's decision to build its own computer hardware, with help from I.B.M., was a direct assault on a market that Intel was counting on for future growth. It is likely that Intel forged the alliance with Apple in an effort to counter the powerful home entertainment and game systems coming from Microsoft and Sony.

[osxisfun]

yah right.

http://blogs.sun.com/roller/page/jon...#an_invitation

[Spliff]

Originally Posted by osxisfun

This part from the article adds to the mix as well:

I must've missed something from the whole Xbox 2/PS 3 hoopla. Why is IBM able to provide Sony and Microsoft with 3.2 Ghz PPC processors, but not for Apple. Is there a huge difference between the console processors and what Apple requires?

I apologize if this has been covered elsewhere.

[sideus]

Xbox/PS3 does not equal G5 processors. Beside, the consoles are not out yet.

[osxisfun]

There is but a difference but the majority of the treatment apple is getting from IBM is that IBM is now making 100 million of the chips for sony - xbox - nifundo and at best, 2 million for apple.

ugly step child syndrome.*

*and that ugly child ain't getting a laptop chip either.

[Superchicken]

I can't wait till tomorrow so this can all be put to rest.

[osxisfun]

My friend just called me, he's doing a demo in a best buy or frys or some such and he said that a guy just walked in off the street and told him that apple is going to intel.

Out of the blue...

WILDFIRE! this rumor is.

[ender2002]

[Don Pickett]

Originally Posted by osxisfun

Is this man our new hero?



Find out why:

http://wiredblogs.tripod.com/cultofmac/

Apparently not:

May 26: Computerworld Today Australia story incorrectly reported that the Intel Pentium D processor and Intel 945 Express Chipset includes unannounced embedded DRM technology.

The Intel Pentium D Processor and the Intel 945 Express Chipset family do not have unannounced embedded DRM technologies.

The DRM technology referred to in a recent report was not a secret or an embedded DRM from Intel. Intel does support various content protection technologies including DTCP-IP technology, which is publicly offered by a number of companies in the industry to enable protected transport of compressed content within a home network.

Many Intel products today support several existing copy protection or content protection technologies. These include support today for things like Macrovision and DTCP-IP, and in the second half of 2005, Intel will deliver an updated graphics driver that will also support additional content protection technologies including COPP, HDCP, CGMS-A, and others. DTCP-IP technology, which is publicly offered by a number of companies in the industry, enables protected transport of compressed content within a home network. While Intel continues to work with the industry to support other content protection technologies, we have not added any unannounced DRM technologies in either the Pentium D processor or the Intel 945 Express Chipset family.

http://www.theinquirer.net/?article=23708

I'm betting Intel starts to make PPC chips.

[osxisfun]

i just read that too at slashdot.

the transitive stuff still applies.

and frankly, how can a guy that looks _that_ dorky not be our hero!!

[el chupacabra]

I thought one of the key perks to OS X was that for the most part only darwin talks to the hardware.
If darwin is the main thing that talks to hardware and its already ported to intel, whats the big deal? Can't you just take whatever hardware specific instructions you need from the windows port of all software and flop it in the new mac?

[BasketofPuppies]

Originally Posted by Superchicken

I can't wait till tomorrow so this can all be put to rest.

Don't be so sure. If Apple announces nothing or announces a more limited parternership with Intel (Intel-powered iPods?) speculation could continue for months.

[sideus]

http://daringfireball.net/2005/06/see_you_intel

Good read.

[Chuckit]

What are you talking about, chupacabra? "The Windows port of all software"? I don't believe all software has been ported to Windows just yet.

[nredman]

[sideus]

Omgwtfbbq!!!!

[el chupacabra]

Originally Posted by Chuckit

What are you talking about, chupacabra? "The Windows port of all software"? I don't believe all software has been ported to Windows just yet.

Not sure if your trying to start an argument... That wasn't the main point. People saying this will bring the mac to its knees. This could only be done if major sofware companies like Adobe etc which have been mentioned in the thread really couldn't make the Intel port (but they have an intel port like most other big companies). Any mac specific software such as a few Ambrosia games isn't going to bring Apple to its knees. Either way although I'm no cs expert the point was even for those mac specific apps it shouldn't be that big a deal since those apps don't talk directly to darwin in general. I know i'm over simplifying, but Darwin seems to be the only thing that would need a major rewrite for this to work.

[CharlesS]

Originally Posted by klinux

I have covered this with you again too. Sure, you would believe Enron and Worldcom committed fruad if Lay and Ebbers themselves said so but that does not mean you'd ignore evidence from WSJ would you? Some people here would.

Hoo boy. Unless there's something somehow illegal about Apple going to Intel for processors, that analogy is beyond terrible.

This report from the WSJ may be correct. Or it may not. Until tomorrow, it's a rumor, because Apple doesn't announce its hardware plans ahead of time, and regardless of whether the source is the Wall Street Journal or MOSR, if they have any valid info, they got it from someone who illegally violated an NDA to give it to them.

That's not fanaticism. That's rationalism. It's not fact until there's an official announcement from Apple. Until then, it's rumor.

[Mediaman_12]

This has got to be Intel manufactured PPC chips.
There is no way after 5+years of 'Mhz & Ghz myth' from Apple Steve can come out on Monday and say "It was just marketing Bull" Intel's chips where faster all along. All those Macworld 'face-off's' rigged, That 40min CPU lecture we got at Macworld a few years back, pointless.

If they do use an X86 derivative, Apples got to to be really careful with the engineering of there motherboard and the rest of the computer, create a 'classic like' environment to run unmodified Windows apps could spell the end of some big apps. Why spend any time porting over to Mac OS to capture that small market-share, when they can get a 'good enough' experience using the windows app and the 'compatibility layer' (it can't be 'seamless' like classic ether, or else say hi to all them Windows based virus').

[hyteckit]

Originally Posted by el chupacabra

I thought one of the key perks to OS X was that for the most part only darwin talks to the hardware.
If darwin is the main thing that talks to hardware and its already ported to intel, whats the big deal? Can't you just take whatever hardware specific instructions you need from the windows port of all software and flop it in the new mac?

Yup. If the application was program using cocoa, then all it requires is a simple recompile to get the application running on Intel x86 chips. That was how the cocoa layer was developed. Darwin is responsible for communicating with the hardware and that already runs on x86 based chips.

Carbon apps on the other handle won't be a simple recompile. Carbon was intended to make it easy for Adobe and other company to port over the existing Classic Mac OS applications over to OS X.

[osxisfun]

That girl in the "4400" ad looks Sith.

[BasketofPuppies]

Originally Posted by CharlesS

and regardless of whether the source is the Wall Street Journal or MOSR, if they have any valid info, they got it from someone who illegally violated an NDA to give it to them.

You're assuming this wasn't intentionally "leaked."

Something this big being reported by three major, credible news outlets is quite likely to have come from a high level executive.

Especially considering ThinkSecret doesn't know anything.

[meelk]

http://meelk.gnbits.com/Intel.jpg

[Don Pickett]

My thots on this:

One of the major promises made to developers with the release of 10.4 was that there would be no major API changes in the foreseeable future. Given this, I can't see any way the switch to x86 is possible. I also don't see the transition to x86 because of the havoc this would visit on Apple's developers. I see a mass exodus of major and minor players away from the platform if this happens, fed up with the fourth major transition in the past decade (68k to PPC to OS X to x86).

Looking at the larger picture, IBM has been slowly getting out of the PC business for the past five years, reinventing themselves as a technology solutions provider. With the sale of their PC division, the PPC processor lines were the only remaining PC resource. Cessation of the PPC lines – accounting for only 2% of their manufacturing – clears them to focus on what they now see as their real business, which is providing enterprise hardware and software and embedded and special purpose processors – console and graphics chips, etc.

Given this prospect, which, if true, we can assume Apple has been aware of for a while, Apple needs to find another chip manufacturer, and who better than Intel? They have the money, the plants, and they can definitely use the low heat and power features of the PPC line. Apple gains considerable marketing oomph – Intel is a household word – and can now sell as many Macs as possible with no worries about constricted CPU supply. Given the famous discounts Intel gives to their marquee vendors, Apple may be able to realize a cost savings per unit on the new, Intel PPC chips, allowing them to either up their margins or cut their costs.

Transitioning the G4-based macs first makes sense, as the G4 is the Achilles heel of Apple's product line because of its low speeds and measly frontside bus. This gives Apple an immediate boost on the low end, allowing (hopefully) Powerbooks to break the 2GHz barrier sooner than later. There is no need to transition the G5-based Macs first, as the G5 still has room to grow (posts on Slashdot from Apple employees state that the next PM speed bump is still G5-based). By the time the G5 machines get transitioned, two years from now, Intel will be up to speed on PPC manufacturing. If this scenario is true, I wouldn't be surprised if IBM and Apple engineers have been working with Intel for a while now.

If the rumors of the Transitive Tech connection are true, this opens an amazing possibility: Apple can sell the ultimate computing Holy Grail – one box which can run all of the major operating systems in use. Imagine buying a Powermac in two years and being able to run OS X, Windows and your favorite Linux distro, plus legacy OS X apps. Imagine the corporate sales. . .

[Kerrigan]

Here is a thought -- Maybe Apple is making an intel computer that doesn't run Mac OS X. The "source familiar with the situation" never said anything about having OS X on Intel. He or she only confirmed that the Mac mini would use an Intel processor, and eventually the higher end models would follow suite.

[fritzair]

Imagine.....
Jobs making more money.
Apple running Intel.
$$$$$
Makes sense.

[osxisfun]

Originally Posted by Kerrigan

Here is a thought -- Maybe Apple is making an intel computer that doesn't run Mac OS X. The "source familiar with the situation" never said anything about having OS X on Intel. He or she only confirmed that the Mac mini would use an Intel processor, and eventually the higher end models would follow suite.

huh?

[Kerrigan]

Originally Posted by osxisfun

huh?

Maybe they are making a windows box.

[Big Mac]

Originally Posted by hyteckit

Yup. If the application was program using cocoa, then all it requires is a simple recompile to get the application running on Intel x86 chips. That was how the cocoa layer was developed. Darwin is responsible for communicating with the hardware and that already runs on x86 based chips.

Carbon apps on the other handle won't be a simple recompile. Carbon was intended to make it easy for Adobe and other company to port over the existing Classic Mac OS applications over to OS X.

Developers attest to the fact that things could not be as simple as straight recompiles. And I think that would apply to both Cocoa and Carbon. The OpenStep frameworks did run on 86k, but it's been a while and Cocoa isn't the same thing as OpenStep. Even if Apple has been maintaining Cocoa for x86, it's not like developers can just recompile and ship. They likely need to go through their code that has been optimized for PPC and optimize instead for 86k. And Programmer over at AppleInsider says the endian issue will still be problematic. Now the only wild card here is Transitive. [i[If[/i] Transitive can deliver (and I seriously doubt their marketing), then Apple can theoretically have a Transitive 86k->PPC layer for us real Mac owners (to translate code compiled for 86k OS X) and a PPC->86k layer for Mactel. I remain highly skeptical that Apple could pull such a transition off, but I'm not going to say it's out of the realm of possibility as long as Transitive can deliver.

[gururafiki]

OK, now I starting to believe it is true now that the NY Times is reporting the story.

[Big Mac]

Originally Posted by Kerrigan

Maybe they are making a windows box.

<Channeling the twit Bill Maher>: "And that's the dumb opinion. . ."

[BasketofPuppies]

Originally Posted by Kerrigan

Maybe they are making a windows box.

Steve hates the Pepsi guy too much for that to happen.

[SoClose]

OK, since everyone else has weighed in with their take on the Apple/Intel rumors, why shouldn't I add my two cents? Here's my thoughts ranked by likelihood:

Apple switches to Pentiums: Not likely. Too many core issues that would cost Apple, developers and users millions of dollars.

Apple and Intel produce PowerPC or PowerPC-like CPUS: Likely. Apple owns significant intellectual property related to the PowerPC design and could utilize Intel's fabs and engineering teams to boost clockspeed, mobile processor production and raise core yields.

Intel produces motherboards and related support chips for Apple: Likely. Until the G5, bus speeds have been abysmal for Macintosh systems. The PowerBook G4s are hamstrung by a pathetically slow bus architecture and memory controller that can't feed the G4 chip enough data to live up to its potential. Intel is very good at making motherboards, something Apple has to do alone right now.

Apple and Intel introduce a joint WiMAX initiative: Quite possible. Apple has a significant history related to wireless technologies. Intel desperately wants to continue to push the Centrino brand and add consumer legitimacy to its WiMAX initiative.

Intel powers the next generation of iPod: The ARM-derived XScale processor could be the platform for Apple to make a revolutionary leap with the iPod platform (they did spin it off into a separate division for a good reason).

Monday will be fascinating, but I'm quite sure that Apple making a wholesale switch to X86 processors isn't in the cards (not that it a small scale shift isn't possible).

[Big Mac]

Originally Posted by SoClose

OK, since everyone else has weighed in with their take on the Apple/Intel rumors, why shouldn't I add my two cents? Here's my thoughts ranked by likelihood:

Apple switches to Pentiums: Not likely. Too many core issues that would cost Apple, developers and users millions of dollars.

Apple and Intel produce PowerPC or PowerPC-like CPUS: Likely. Apple owns significant intellectual property related to the PowerPC design and could utilize Intel's fabs and engineering teams to boost clockspeed, mobile processor production and raise core yields.

Intel produces motherboards and related support chips for Apple: Likely. Until the G5, bus speeds have been abysmal for Macintosh systems. The PowerBook G4s are hamstrung by a pathetically slow bus architecture and memory controller that can't feed the G4 chip enough data to live up to its potential. Intel is very good at making motherboards, something Apple has to do alone right now.

Apple and Intel introduce a joint WiMAX initiative: Quite possible. Apple has a significant history related to wireless technologies. Intel desperately wants to continue to push the Centrino brand and add consumer legitimacy to its WiMAX initiative.

Intel powers the next generation of iPod: The ARM-derived XScale processor could be the platform for Apple to make a revolutionary leap with the iPod platform (they did spin it off into a separate division for a good reason).

Monday will be fascinating, but I'm quite sure that Apple making a wholesale switch to X86 processors isn't in the cards (not that it a small scale shift isn't possible).

Yeah, but as ridiculous as it sounds, the WSJ and NY Times report the 86k transition as fact. Either they're all wrong, or it's time to prepare for the end of the world.

[osxisfun]

Good post soclose.

As "idrivex" said (knows inside info?) said:

"Seriously, it won't be that big a deal. Promise."

That (to me) leads to intel fabbing the ppc and as I posted before steve could do this on stage:

"Hey. we have a new partner. Intel is going to start making our powerpc chips starting next year. No biggie. Its going to be the same powerpc you know and love only better."

then he could go onto the next bullet point of his keynote

stock market happy.
developers happy.
users happy.
HW sales over the next 2 years - same or better.


and one more thing:

Intel is going to help us make our laptop chips.

[Big Mac]

In this eWeek article they report an Apple spokeswoman characterizing the WSJ article as rumor and speculation.

[Chuckit]

Originally Posted by Kerrigan

Maybe they are making a windows box.

Thank you. Now that we've actually got a reasonable theory as to what Apple would be doing with Intel, I was worried this thread might just run out of ridiculous.

[Kerrigan]

Originally Posted by Chuckit

Thank you. Now that we've actually got a reasonable theory as to what Apple would be doing with Intel, I was worried this thread might just run out of ridiculous.

It would be the smart thing for them to do. Everyone knows what a Mac mini is right now, but they are not selling well because nobody wants to buy a computer running OS X.

[Don Pickett]

I have to say: if Apple does go x86, I will keep my newly purchased G5 forever. It will be the last in a long line.

[Chuckit]

Originally Posted by Kerrigan

It would be the smart thing for them to do. Everyone knows what a Mac mini is right now, but they are not selling well because nobody wants to buy a computer running OS X.

"Ladies and gentlemen, just a few months ago, we introduced you to Mac OS X Tiger and told you it was the most advanced operating system in the world, years ahead of Microsoft Windows. Never mind all that; I was really drunk at the time. But how about this nice new Windows PC that's still more expensive than comparable Dells?"

[Kerrigan]

Originally Posted by Chuckit

"Ladies and gentlemen, just a few months ago, we introduced you to Mac OS X Tiger and told you it was the most advanced operating system in the world, years ahead of Microsoft Windows. Never mind all that; I was really drunk at the time. But how about this nice new Windows PC that's still more expensive than comparable Dells?"

Sounds like a more viable idea to me than dragging everyone through a processor switch and losing the very few customers who are still hanging onto Apple. It took nearly 5 YEARS just to have an OS that was as responsive as Mac OS 9, now everything has settled down, and they want to change everything up again.

The consensus is that a switch to Intel is suicide for Apple, that Adobe and the rest of the major developers will not want to go along. Sounds to me like Mac OS X is damned if it switches to Intel and damned if it stays with IBM. It might just be time for Steve to experiment with Windows based computers.

[klinux]

Originally Posted by Big Mac

In this eWeek article they report an Apple spokeswoman characterizing the WSJ article as rumor and speculation.

Duh, what did you expect the spokesperson to say? And did you read the actual article rather than a retold version?

The important thing regarding that spokeswoman is that she did not deny the Apple and INtel executives meeting.

[Don Pickett]

Originally Posted by Kerrigan

The consensus is that a switch to Intel is suicide for Apple, that Adobe and the rest of the major developers will not want to go along. Sounds to me like Mac OS X is damned if it switches to Intel and damned if it stays with IBM. It might just be time for Steve to experiment with Windows based computers.

The consensus is that a switch to x86 could be very difficult. A switch to Intel manufactured PPC chips is a good thing.

[hyteckit]

Originally Posted by Big Mac

Developers attest to the fact that things could not be as simple as straight recompiles. And I think that would apply to both Cocoa and Carbon. The OpenStep frameworks did run on 86k, but it's been a while and Cocoa isn't the same thing as OpenStep. Even if Apple has been maintaining Cocoa for x86, it's not like developers can just recompile and ship. They likely need to go through their code that has been optimized for PPC and optimize instead for 86k. And Programmer over at AppleInsider says the endian issue will still be problematic. Now the only wild card here is Transitive. [i[If[/i] Transitive can deliver (and I seriously doubt their marketing), then Apple can theoretically have a Transitive 86k->PPC layer for us real Mac owners (to translate code compiled for 86k OS X) and a PPC->86k layer for Mactel. I remain highly skeptical that Apple could pull such a transition off, but I'm not going to say it's out of the realm of possibility as long as Transitive can deliver.

Big-endian, little-endian. It doesn't matter. If the application was coded using the cocoa framework, a simple recompile for the Intel chip would solve the big-endian, little-endian. Darwin is what's responsible for communicating with the hardware, which includes the memory. Darwin already runs on x86 chips, so big-endian and little-endian is not an issue.

http://developer.apple.com/documenta...I_BOOK.1f.html

http://support.microsoft.com/default...b;EN-US;102025

[nforcer]

I hope these reports of a switch to x86 end up being true. I hope it's a "complete" switch (ie, we could install a new version of OS X on any modern x86 machine), not "we've gone with x86, but you still need to buy from us because you will require a special ROM chip that only we sell" type switch. I don't care about the whole "PPC is better, x86 is a CISC POS" debate. PPC may be better, but look at the current situation. The G5 developments over the last 2 years are still far from where we were promised. The chip is a hot chip - too hot to do laptop development 2 years after its release, apparently. And I would argue most importantly are the problems with supply.

With the first dual 2GHz Powermacs full scale shipping to customers took months after the official release date. Some 2.5GHz Powermacs took up to 6 months to get into customers' hands. Even then major problems and supply issues persisted for months. Almost a year later we're at 2.7GHz thanks to the continued wonders of liquid cooling. How does Apple expect to grab more marketshare when they have problems supplying the 3% they currently have?

The PPC might have a good future since all the next-gen consoles using some kind of PPC thing, but it might be too long to wait and take a risk on. From what I have read at Ars Technica, Cell and whatever MS is using are designed to be great for graphics but are not so great for general purpose computing. Apple falls under the latter category and needs to do better than they have been doing and preferably soon. The situation with Motorola hurt them badly and the current situation with IBM feels like the same. Remember the MHz Myth and all of that marketing? At the current rate we'll be resurrecting the MHz Myth arguments next year. While a move to x86 would more or less be flushing a lot of marketing down the toilet, in the long term Apple wouldn't need to spend any more resources fighting Intel and co: they could just do business with them.

The "I won't pay for all of my apps again" argument is a poor one. Apple going to x86 would not force you to upgrade or pay anything. You can continue to use what you have paid for. Furthermore the statement implies existing updates for apps are free and that's far from the case. Don't we go through this "why should I have to pay" debate with every OS X update (sans 10.1 because that was free)?

"But the change would mean I wouldn't be able to use SoundApp anymore!". Sucks that you will no longer be able to use a dead app, but you can continue to use whatever apps you like on whatever setup you currently have.

"But developers will have to redo a lot and we might lose the support of Adobe and the companies they have purchased!" Last I checked Adobe and some other big companies derive a nice profit from the Mac market despite its size. Even if a switch gives them a lot of work to do, they have to realize that with a complete switch the Mac OS could potentially be running on the other 95% of computers. That creates a much larger market opportunity than what exists now. The barrier for potential switchers would be lowered a lot by letting them use hardware they already have.

Yes, there are a lot of immediate downsides. All of the code that has been optimized for PPC would be gone. All of the code that uses Altivec would be gone. A lot of technical things would have to change and it would require a lot of work in some cases. But doing extra work to keep existing applications running is not really different that it is now: every OS X update has broken a bunch of existing apps in some way. It's the price of progress and in the long term it would pay off.

My overall argument is that the underlying hardware should be irrelevant to users: it's the OS, interface, and applications that distinguish the Mac platform. With a move to x86 Apple could continue to sell hardware with pretty cases or complete setups; we would just have the choice to buy from another vendor and more options. Being able to have Windows on one partition for gaming and OS X on another partition for everything else would be great. Having access to better video cards and other parts would also be a plus. There would be a lot of short term pains but even more long term gains and potential.

[Oneota]

I dunno...I just got my badge for WWDC, and all the big-ass banners were already unfurled, leading me to think there's nothing big in the works (last year, all the banners with specific references to new features were covered until after the keynote - only the big Tiger CD ones were viewable).

[osxisfun]

15 hrs you martian beactches....