Quick Transition or Slow -- Thoughts?
The Mac Rumor echo chamber seems to have concluded that Apple's transition to Intel will be drawn out over a period of up to two years. Plus it's commonly assumed that PowerBooks and iMacs will be first, and workstations (PowerMacs) will be transitioned much later.
As far as I can tell, there's no evidence from Apple for these notions. Some of this speculation seems just 'little white lies' that people tell themselves so they don't feel scared about buying a new PPC.
My feeling is that a drawn out, "consumer/laptop first" transition would be a big mistake. Apple's biggest problem is convincing ALL developers to support Intel, and that's a much bigger issue in the professional/desktop space than consumer.
Furthermore, having both Intel and PPC on the shelves at the same time will inevitably cause confusion, and I strongly believe that it would behoove Apple to move fairly quickly and have a *complete* lineup of Intel machines as soon as possible. If I were Steve Jobs, this is the approach I would take:
A) Have some sort of Intel desktop machine available from the first day. This might not have to be a complete PowerMac replacement -- perhaps something simple to engineer and cheaper like a Pentium-D desktop. Something like the ADC machines, but generally available. This is critical for developers and IT departments to start testing.
Have a totally redesigned Intel PowerBooks (new enclosures) available from the first day, preferably with the nextgen Pentium-M chips. This is critical to get the "fashion leaders" on the Intel bandwagon.
C) Move all consumer machines in one fell swoop within the first 3-9 months. Just pick a date and cut over the iBook/iMac/Mini all at once. There's no sense in having a Intel iBook sitting next to a PPC Mini at the Apple Store. This may mean making design compromises on iBook/Mini.
D) Finally, create a true PowerMac workstation replacement, probably 2x2x Xeon based. This is a declining segment for Apple, and the hardcore Audio/Video editing crowd will be the most conservative about the switch. XServe would use a similar platform.
E) Keep legacy PPC PowerMacs and PowerBook 15s in the catalog indefinitely as long as there is demand. Make it clear to pro customers that PPC won't disapppear for a long while.
Counter Arguments: People have pointed out how the 68K->PPC conversion took years to complete. The truth was that Apple used CPU compatibility to segment their market pro vs consumer. It was a mistake then with 15% marketshare, and would be an even bigger mistake now with 3%.
There is also the argument that with Rosetta and XCode, CPU compatiblity won't be a big deal. I don't buy this. It's too confusing for consumers. Plus Apple's developers are going to stongly want only one platform to support.
As far as I can tell, there's no evidence from Apple for these notions. Some of this speculation seems just 'little white lies' that people tell themselves so they don't feel scared about buying a new PPC.
My feeling is that a drawn out, "consumer/laptop first" transition would be a big mistake. Apple's biggest problem is convincing ALL developers to support Intel, and that's a much bigger issue in the professional/desktop space than consumer.
Furthermore, having both Intel and PPC on the shelves at the same time will inevitably cause confusion, and I strongly believe that it would behoove Apple to move fairly quickly and have a *complete* lineup of Intel machines as soon as possible. If I were Steve Jobs, this is the approach I would take:
A) Have some sort of Intel desktop machine available from the first day. This might not have to be a complete PowerMac replacement -- perhaps something simple to engineer and cheaper like a Pentium-D desktop. Something like the ADC machines, but generally available. This is critical for developers and IT departments to start testing.
Have a totally redesigned Intel PowerBooks (new enclosures) available from the first day, preferably with the nextgen Pentium-M chips. This is critical to get the "fashion leaders" on the Intel bandwagon.C) Move all consumer machines in one fell swoop within the first 3-9 months. Just pick a date and cut over the iBook/iMac/Mini all at once. There's no sense in having a Intel iBook sitting next to a PPC Mini at the Apple Store. This may mean making design compromises on iBook/Mini.
D) Finally, create a true PowerMac workstation replacement, probably 2x2x Xeon based. This is a declining segment for Apple, and the hardcore Audio/Video editing crowd will be the most conservative about the switch. XServe would use a similar platform.
E) Keep legacy PPC PowerMacs and PowerBook 15s in the catalog indefinitely as long as there is demand. Make it clear to pro customers that PPC won't disapppear for a long while.
Counter Arguments: People have pointed out how the 68K->PPC conversion took years to complete. The truth was that Apple used CPU compatibility to segment their market pro vs consumer. It was a mistake then with 15% marketshare, and would be an even bigger mistake now with 3%.
There is also the argument that with Rosetta and XCode, CPU compatiblity won't be a big deal. I don't buy this. It's too confusing for consumers. Plus Apple's developers are going to stongly want only one platform to support.
Comments
As for the nature of the transition, that's been reported by journalists who have talked to Apple people, and it just makes sense.
Originally posted by BRussell
Well Jobs laid out their timeline in his WWDC keynote: It will have begun by June 2006, will be mostly done by June 2007, and will be done by the end of 2007. I'd say that's a pretty quick transition, given what they're doing.
As for the nature of the transition, that's been reported by journalists who have talked to Apple people, and it just makes sense.
The key verbiage from the mouth of Steve Jobs is that the transition will begin next year, and that "by this time next year" (June 2006)...so I take that to mean anytime between January and June we will start to see Intel Mac rolling out. What will be first seems likely to be in the portable range, however, it's anyone guess at this point. The Paris show, a month from today, will give us an idea of what we may expect at the MacWorld show in January, and probably an idea of when we can expect the Intel roll out to begin.
I think it's exciting times for Apple, and I look forward to using the new Macs with Intel processors.
The transition is only an issue for developers, and is mostly a good news story at that. xCode really does make the transition for most programs super-simple. Being on Intel means that game ports (eventually) will happen faster. Major players like Microsoft, Adobe and others are already on board.
For consumers, and professionals, they're still getting a Mac. Having used one of the development boxes, I can tell you that there's nothing confusing here. Everything works exactly the same.
A Mac is a Mac is a Mac.
Plus, by doing this everything will be in XCode and better optimized for the mac platform.
(Some developers will drag their feet on XCode as long as possible, Rosetta may not work or work well for everyone, etc.)
How is this confusing for the consumer? Last I checked people bought computers not processors. They buy software for operating systems not motherboard architechtures.
Right -- and that's why Apple should transition as quickly as possible. Otherwise, like it or not, the consumer would be facing a confusing decision about processors and motherboards.
(Even though Apple tends towards selling a "lifestyle", they are still a computer company, and the specs are still sitting there right in front of people.)
Originally posted by IntlHarvester
Composer -- I admire your optimism, but you've clearly drunk the Kool-Aid when it comes to the software situation. It will not be as simple as you say, in particular with niche titles and the existing game library.
(Some developers will drag their feet on XCode as long as possible, Rosetta may not work or work well for everyone, etc.)
Right -- and that's why Apple should transition as quickly as possible. Otherwise, like it or not, the consumer would be facing a confusing decision about processors and motherboards.
(Even though Apple tends towards selling a "lifestyle", they are still a computer company, and the specs are still sitting there right in front of people.)
There is no need to disparage Composer. If you want to be confused about the transition, that is your right. It doesn't mean that the rest of us will be.
Originally posted by Mr. Me
There is no need to disparage Composer. If you want to be confused about the transition, that is your right. It doesn't mean that the rest of us will be.
So, do you have an opinion on the topic, or are you too busy goosestepping?
Originally posted by IntlHarvester
So, do you have an opinion on the topic, or are you too busy goosestepping?
So you returned to this forum to insult everyone?
As for the topic at hand, the point that you seem to to miss is that MacOS X's developers have been doing multiplatform development for more than a decade. It has worked very well all this time. There is no reason to believe that it won't continue to do so. Without any evidence to support your view, you want us to be afraid that somehow it will all go awry.
As for Rosetta, you assume that Apple's commercial release of MacOS X/Intel will include only a half-@$$ implementation of the technology? Will Rosetta be perfect? Of course not. Will most users see its imperfections? Probably not.
Originally posted by IntlHarvester
Composer -- I admire your optimism, but you've clearly drunk the Kool-Aid when it comes to the software situation. It will not be as simple as you say, in particular with niche titles and the existing game library. . .
Well then, the developers I know must have drunk the Kool Aid too, since they see little or no problem porting most of their applications. You are correct that some titles will need more work, especially if they use MMX/SSE or whatever it's called, since that will take some rewriting. Even then they have an answer. It needs to be rewritten to make a Windows XP version anyway.
Here is a list of problems that will come up:
(1) People with Intel Macs will buy older versions of Mac software which are still in stores, and it won't work. They will be very unhappy about this.
(2) Developers who are using CodeWarrior will say heck with it, the transition will be far too difficult to be worth it, and keep doing what they're doing, reasoning that most Macs will still be PPC for a while. Many will simply stop developing for the Mac. Some packages will be labelled "PowerPC Only"; some won't. Consumers will buy them, and some will be unhappy.
(3) Endian issues will arise in many packages, meaning that the Intel or dual-binary versions will be delated for months or even years, or possibly never come out at all.
(4) Some developers will be attracted by the Intel CPU and produce Intel-only versions of their Windows software, avoiding the endian issues. There may be labels that say "Intel Processor Only". Consumers won't see them. Many will be unhappy.
This is not pessimism. This is facing the facts. Anyone who cavalierly dismisses the processor and endian issues shows he has no experience in programming, and doesn't know what he's talking about.
To say "A Mac is a Mac is a Mac" is wishful thinking. It's not true today, it's never been true in the past (unless you go back to 1984), and it will certainly not be true when this transition is underway.
As to the original topic, BRussell is right: this is a quick transition. To us sitting in our armchairs, two years may seem like a long time, but there are a lot of machines to transition. Given all the design, production and inventory issues, this is a very aggressive schedule. There will be some grim aspects to it, so t'is better t'were done quickly.
Let's try to get back on track by restating things:
There's going to be a period of up to one year when PPC Macs are sitting next the shelf next to Intel Macs. Most buyers are going to have to make an educated decision. Sure. there's a certain market that will buy anything with an Apple logo on it just to run iTunes, but even if the software situation is perfect, everyone else is going to have to consider running Rosetta versus paying upgrade fees versus buying PPC.
I don't believe Apple wants to put it's customers in a position where they must juggle such technical decisions. (I'm trying to avoid the word 'confusing' here...) So, I think the best route is a hard cutover. As soon as MS Office and Photoshop hit the market, just remove the CPU decision all together and go 100% Intel across the line, all at once.
(At the same time, provide a clear roadmap for "legacy" support so that IT Managers and the like can still buy PPC from the back of the website while they plan for the conversion.)
Finally, I think tacojohn said it best -- some people want a drawn-out transition, not because it would be the most ideal strategy for Apple, but because they want to protect their own investment. I wouldn't worry about that -- G5 PowerMacs will still have a long usability cycle, and most G4s were going to be a natural victim of "Gates' Law" (Every 18 months the speed of software halves) sooner or later anyway.
Originally posted by IntlHarvester
Mr Me -- I am making no predictions that things will "go awry", this is not a FUD exercise. Apple has all the right pieces in place to pull this off. However, the idea that the transition will be totally invisible to consumers and pros is not reasonable at all.
....
You say that "this is not a FUD exercise," but then you spread more FUD. It would be one thing if we had never before gone through a processor transition before, but we have. In fact, the 680x0 to PPC transition is not yet complete for users who must run Classic apps. Even though the original MacOS was never meant to see a processor family transition, Apple took it through the transition beautifully. Now MacOS X is based on OpenSTEP, which has run on multiple processors for more than a decade. It has been that way ever since NeXTstep became OpenSTEP. As for MacOS X, the transition is not the OS to the processor, but the user base to the processor. There has never been a time when MacOS X did not run on Intel-based computers.
You may deny it all you want, but you are as transparent as the air we breathe. You are spreading FUD.
This is definitely a lofty goal. But as of yet, we have no indication to think it a faulty precedent.
Originally posted by IntlHarvester
There is also the argument that with Rosetta and XCode, CPU compatiblity won't be a big deal. I don't buy this. It's too confusing for consumers. Plus Apple's developers are going to stongly want only one platform to support.
If the developers wanting only one platform to support was a show stopper, Apple would not have made the jump. No matter whether Apple spreads the transition over two years or two weeks, we will be living in a multi platform enviroment for several years yet. Macs last a long time. All the machines in the channel and all the refurbs/surplus machines (which will lag new machines for quite a few months) will be landing in the hands of the user base for quite some time after . It's going to be several years before most developers will feel safe dropping the PowerPC.
There will be lots of grumpiness all around. Developers will want to do the Intel Switch at a Version upgrade and charge quite a bit for it. Some Applications won't like Rosetta. People won't read the box and will have problems returning software that they've opened which does not run on their machine. Some users and developers will decide to get off the train. We will hear about all of this in painful detail in comments to Macintouch and the message boards.
I could see Apple introducing a Mini G5. They need to ignite some excitement to keep boxes moving until the Intel Macs are ready for Prime Time.
And I wonder how long it's going to take M$ to follow Apple's lead and use trusted computing to block OS and even application piracy.
Originally posted by Mr. Me
As for MacOS X, the transition is not the OS to the processor, but the user base to the processor.
Great, I'm glad you've finally figured out what the topic is.
You may deny it all you want, but you are as transparent as the air we breathe. You are spreading FUD.
On this thread, you have been nothing but an argumentative zealot who has contributed no opinion on the subject at hand (model release timing). Advocating a "hard cutover" is not even remotely close to Fear/Uncertainty/Doubt.
But hey, if you are scared of FUD, lay it at the feet of Apple -- they're the one who announced the obsolescence of their entire product line with no detailed roadmap. If Jobs wanted to stop speculation on the rumor boards, it would be quite simple to do so.
Originally posted by BobtheTomato
If the developers wanting only one platform to support was a show stopper, Apple would not have made the jump.
I think rather they only one one platform to support, and they want it to be x86. But they understand that means a transition period. At least the larger developers that maintain a Mac/Windows code base.
No matter whether Apple spreads the transition over two years or two weeks, we will be living in a multi platform enviroment for several years yet. Macs last a long time. All the machines in the channel and all the refurbs/surplus machines (which will lag new machines for quite a few months) will be landing in the hands of the user base for quite some time after . It's going to be several years before most developers will feel safe dropping the PowerPC.
I agree -- it's really a question of whether PPC support is mainstream for 3 more years (1 major release) or 5 more years (2) from the first day. Both are fine for the current installed base. For applications with good file-format compatibility, such as MS Office, they can move more quickly wihout affecting the user base much (just buy the old version).
Speaking of Office, I hope that Rosetta meets its goal of running it well, because that will make the transition $200 cheaper for a lot of users (including myself).
I could see Apple introducing a Mini G5. They need to ignite some excitement to keep boxes moving until the Intel Macs are ready for Prime Time.
Why wouldn't a Fast, game-capable Intel Mac be exciting enough for the Mini crowd? Putting a new PPC Mac on the shelf next to an Intel one just defies the "a Mac is a Mac" idea, which makes it seem to me quite unlikely.
Originally posted by cubist
. . . Developers who are using CodeWarrior will say heck with it, the transition will be far too difficult to be worth it, and keep doing what they're doing, reasoning that most Macs will still be PPC for a while. . . . Some developers will be attracted by the Intel CPU and produce Intel-only versions of their Windows software, avoiding the endian issues. . .
Don't you suppose that CodeWarrior would update their development kit before long? It is eithet update or get out of the business and leave it all to Apple tools. . . . Also since Apple has such a good set of tools that produces code for both PPC and Intel processors, why on earth would a developer wish to produce software that only runs on Intel Macs and miss over half the market for his or her product? It makes little sense.
In short, every party involved in the transition process has to think for themselves.
Unless there is a Sossaman surprise with 64-bit support, it's going to be sometime in 2007 when Conroe comes. The transition could be complete by WWDC2007 (1H 07) if Conroe comes in Q2 07.
Yonah, Sossaman in Q1 06 means laptops, Mac Minis and eMacs could be Mac/Intel machines in by WWDC06.
Merom in Q3 06 and Conroe, Woodcrest sometime later, mean Power Macs and Xserves could be Mac/Intels by Q4 06 or by WWDC07 depending on when Conroe comes.