Will Vista Be the Last Operating System Microsoft Produces?
I don’t think Microsoft will ship another Operating System after Vista launches. I believe that a combination of technical difficulties and changing markets will prevent it from creating a product that is relevant in the market. Consider this, if the latest shipping dates are to be believed, it will have taken Microsoft over six years to get Windows Vista out the door and to its consumers. And based on past events, it is safe to assume that Vista will require at least one service pack before it is truly ready for use. Of course, factoring in the normal Microsoft delays for producing patches, such a comprehensive service pack will probably take another year before it can be released to users. That would mean that it will have taken Microsoft 7+ years to make a usable operating system.
Now consider how long it could take Microsoft to produce Vista’s successor. If the added complexity of this new OS increases the development time by only 25% (not an unreasonable figure) of what it took to make Vista, then it will have been in development for almost 8 years. That means if Vista comes out in 2007, it won’t be replaced until 2015. To put that into perspective, if Apple continues on with its release cycle of OS X (and factoring in increases in development time) they could, counting Leopard, release 4 to 5 new operating systems by the time Microsoft releases one.
But keeping up with Apple won’t be Microsoft’s biggest concern. What will prevent Microsoft from releasing another OS is the changing market. For Vista’s successor to have a hope of selling, the company has to assume that no fundamental shifts in technology will occur for almost a decade! That seems, overly optimistic at best. With Google threatening to release a web-based OS, and Apple potentially using virtualization to run all Windows applications, Microsoft might find that by the time it can cobble something together, it no longer has a market interested in its product.
Microsoft will find itself in this position (or one like it) all too soon, and it has no one but itself to blame. Here are the two biggest factors that are slowly killing Microsoft from within.
Code base
The amount of code that makes up Windows has simply become too large to work with. Now, you can blame this on anything you want (backwards compatibiliy would be high on my list), but ultimately the cause doesn’t matter. What matters is that building new features has become impossible, and debugging this mess has become impossible + 1. This was most clearly witnessed when Bill Gates got up onstage and informed his eager audience that the codebase for Vista had become so large and tangled that they simply had to throw it all away and start over from a point they knew was stable. Guess what? That problem isn’t going to go away by throwing another service pack at it. With each version of Windows released the amount of code grows and the strain gets greater. However, the amount of code isn’t the only problem here. The structure of the OS itself is fundamentally flawed. There are too many antiquated ideas (drive letters, the registry, etc.) and constraining bounds (NTFS) to allow for anymore growth. A drastic rewrite is the only way to solve this problem. The only real question Microsoft needs to ask is how much should we rewrite?
Management
The last few years has seen a flurry of restructuring at Microsoft. Key people have left (most noticably for Google) and even loyal employees who still believe the hype have begun to criticize management and air their grievances on personal blogs. The leadership of Microsoft has failed miserably and Vista is only the beginning in what looks to be an impressive series of embarrassments. It is time for a change. If Microsoft still hopes to be in the OS market a decade from now then those changes can’t come soon enough.
del.icio.us
Digg
Facebook
Slashdot
StumbleUpon
Comments
James, I have no proof of this, but I think Microsoft bit the bullet two years ago. It’s the reason that Longhorn (now Vista) was delayed. Longhorn’s complexity grew so great that it became unmanageable. So, what does Microsoft do when its plans fail? It looked to Research South in Cupertino: that is, it copied Apple again.
Do you know why Pink and Taligent didn’t work? They were trying to create an integrated OS. And after a while, that becomes impossible. Yes, legacy hardware and software was part of the problem, but they still had that problem to solve with Copeland (Mac OS X.) The solution that Next provided to Apple was to build the OS in Modules on layers with functions which communicated with each other. That way you could change a part of a layer (a module such as Quicktime, say) without breaking everything on a higher layer.
Mac OS X has the mach kernel that talks to the hardware. Above that is the Darwin Open Source BSD Unix layer. Above that is the Cocoa Application Programing Instruction Set and the Carbon API’s which allowed Apple to run updated legacy Mac OS 9 software. And if you believe I, Cringley, the Microsoft Windows XP API’s will be added in here in Leopard 10.5. Above that is Aqua, Quicktime, CUPS and a host of other operating system applications. And finally, you get up to the user applications which you run.
Building an OS in modules allows you to change the internals of a module or the individual API’s without breaking everything above. What I understand is that Apple only recently completed it’s plans for the Cocoa and Carbon API’s and has declared them mature, yet the Mac OS still ran for six years, just not as well as now.
I suspect that Microsoft had to copy that modular system. This means that Vista will break many legacy applications and hardware the way that Mac OS X 10.0 did. Vista will not run on some computers that are being built now; that is the point of the Vista Capable and Vista Ready labels. Computers older than a few years will be forced to stay with Windows XP.
If Microsoft did bite the bullet, then it will take the part of it’s integrated OS which works acceptably and will increasingly be moving more of its subsystems to a modular system. Upgrades may be as fast and furious as the Development of Mac OS X was.
This will not be convenient for either Microsoft or its users. Many current users will not be moving to Vista for some time. Partly, this is because there are many older versions of Microsoft Windows still being used and people often wait until their computer breaks down completely or some software that they need no longer functions. Then, they will be forced to upgrade to a new computer with Vista and, most likely, will have to buy new software.
This presents Apple with an opportunity since the compatibility issue no longer matters. A consumer could buy a new Dell or HP computer with Vista or a Macintosh that runs both Macintosh and Windows XP programs transparently. Apple will not have to become a Microsoft OEM (as it has promised), because it already has the right to copy the Windows API’s from its deal with Microsoft to share technology seven years ago. Since the Windows XP programs will run within Mac OS X’s more secure shell there will be few differences in security, convenience or looks. The Applications will just work. And they will work on the Mac faster and better than they ever did in Windows.
Meanwhile, most of the developing world will be moving to Linux, because of the $50 to $100 that Microsoft charges OEM’s. As the cost of computer components declines and more of the components are moved onto the processor chip, the Microsoft tax will eventually become larger than the cost of a low end computer. China and India will be pouring billions of dollars into making the Linux desktop work. There will be enough problems with the Linux desktop to persuade anyone who can afford to do so to switch to a Mac or a Windows machine. But when most of the world can only afford a $100 computer, they can’t afford another $50 to $100 in Microsoft tax.
Apple will try to take over the top 25% of the computer market based on ease of use, flexibility and speed. Microsoft will be forced to play catchup while it’s market share declines. Competition will be fast and furious. Many changes are coming in hardware as each component part of a computer gets it’s own controlling processor chip. It is unclear who will win this game, but there are exciting times ahead.
I can remember when IBM’s 360 was the king of the OS world. Kinda shows that things can change over time.
that is the problem that MS faces today. They have made a fortune holding onto backward compatibility, but it is an albatross hanging around their neck these days. So what can they do? First is to deliver Vista and continue to upgrade it with new features to hold them over for a while.
Then they need to look at a new OS. Apple forced them to move from DOS to Windows and now the albatross is going to force them to look at what that new OS should be. Actually start with a clean sheet of paper and get the best brains to work on it.
IBM did that in the 50’s - got their best brains to come up with what would be the ideal computer for the year 2000. One thing that came out of that effort was the System 38, which had incredible features for it’s day.
Oh, and keep the accountants out of the effort. MS needs these guys and gals to work out what would be the design for the future - there is no room for the bean counters trying to figure out how many versions it should have and how they can obtain ongoing revenues. MS has the talent to get the job done - if BG and the Bean Counters can stay out of their way.
To anyone who might take the uninformed conjectures above as truth:
Today’s Vista began life in early 2005, when MS started over from scratch. That is, started over from the latest stable code base (Windows Server 2003 SP1) At that time, the idea was to tackle the problem of the over-complex code. Now, you can’t turn around in Vista without finding a core sub-system that’s been completely rewritten. Video, audio, networking, you name it. They’ve been redone from the ground up with the future in mind. So you see it’s quite the opposite. After this initial hump, MS will be poised to crank out new releases.
Some people mention the sidebar vs the dashboard. The sidebar will be available to the masses after the dashboard, but does that mean the idea was stolen? No, in fact you can see the sidebar in Longhorn builds that were made way before OS X Tiger. Several other features were planned for Vista all along, as can be seen from old press releases / articles. These too are claimed to have been stolen from Apple, just because they are out in the current version of OS X.
On another note, the ‘they stole from OS X’ argument is a dead horse. Windows, OS X, and Linux ALL owe some thanks to other OSes.
hdjdnda wrote:
“Today’s Vista began life in early 2005, when MS started over from scratch. That is, started over from the latest stable code base (Windows Server 2003 SP1) At that time, the idea was to tackle the problem of the over-complex code. Now, you can’t turn around in Vista without finding a core sub-system that’s been completely rewritten.”
Yes, that is exactly what I had read. And that is what I based my speculations on.
“On another note, the ‘they stole from OS X’ argument is a dead horse. Windows, OS X, and Linux ALL owe some thanks to other OSes.”
Stole is the wrong word because Microsoft had every right to use Apple’s technology from their deal seven years ago. It was that Microsoft had a blunder on it’s hands, longhorn wasn’t working right and they went with technology that they knew worked. I meant no disrespect to Microsoft here. They were smart to do it.
“Video, audio, networking, you name it. They’ve been redone from the ground up with the future in mind. So you see it’s quite the opposite. After this initial hump, MS will be poised to crank out new releases.”
Yes, that is my point exactly. But, most of those releases will be in providing the fundamentals which will run well together, just like what happened with Mac OS X. It wasn’t until 10.2 that Mac OS X was an acceptable desktop operating system, not until 10.3 until it became superior to Mac OS 9. Microsoft has lots of hard work ahead just to tread water.
The problem that I see is that Microsoft might have started too late. MS has tons of money to throw at the problem and plenty of engineers. But, building in quality is going to hard. Their previous slap-dash methods won’t help them here. The Mac OS is a standard that they are going to be compared harshly against. They had better get it right. And I don’t know if two years of hard work is enough. A lot of the code is going be a kluge underneath. It will work, but they’re going have to go back and fix it by providing the same kind of robust support structure that Apple already has.
“1991: Mac OS 7.
1997: Mac OS 8.”
WHAT DOES THIS HAVE TO DO WITH ANYTHING?? Yes, the caps are diliberate. Yes I appricate that you felt the need to share your point of view but, hey, we’re talking about the year 2006. We’re talking about XP-->Vista vs. Mac OS X. Please, for the love of logic, stay on topic. Seriously. That is just poor argumentation. It is completely irrelevant and, furthermore, completely glosses over the releases between System 7 and 8 which were significant releases and not just ‘service packs’.
Thank you.
Signed,
Someone who cares about actually engaging with the topic at hand instead of using comments as sounding boards for one’s ridiculous prejudices and poor reasoning.
I never heard of this site until Google News started to headline its articles. What a disappointment it is… I thought the article (judging by the title) was going to have some interesting points and insights. Instead it just has wishes and hopes of an ankle grabbing fan boy.
Look man, you need to read the writing on the wall. Apple is getting ready to become a premium hardware supplier not take on MS. They went to the intel chip because they recognize nobody does HARDWARE like them (e.g. iPod), but their software just isn’t accepted by the masses. For example, Apple’s OS just recently passed the amount of current usage of WINDOWS 98! LOL. Apple is a threat to Dell not Microsoft. http://marketshare.hitslink.com/report.aspx?qprid=5
I hate to see such a diehard “groupie” setting themselves up to be blind sided like this… its kind of sad really. Let me do a public service and help you understand why Apple isn’t going to beat MS, its called xcode. It’s a good tool for OSX development, but don’t mistake it as a good development tool. You are dead wrong thinking MS is stupid for bloating code to support backwards compatibility. Apple should have bloated their code and supported OS9; they lost a lot of Apple software with the OSX upgrade.
Lesson 1: Developers and software decide the success of an OS. Windows has it, Apple doesn’t. The reason MS has such a large market share is because of the tools they provide, and have provided for well over two decades. You can’t just replace that with a new program and 5 years. The reality is Apple doesn’t have it anyways. Windows has a legion of developers and software grandfathered in, they are comfortable with Windows, and won’t just change over to OSX because you keep kicking and screaming “apple is great, apple is great, all hail Steve”.
Businesses for the most part write software exclusively for Windows, and in many cases web-based application are exclusively for IE. This is a fact, and its the reality, something you should visit more often. Honestly, it makes good sense for a Business to write for Windows and not Apple OS (Apple has less then 5% saturation). It’s about MONEY in the real world (reality), decisions certainly are made with your emotional agenda.
It costs lots of money to write a program for multiple operating systems and to support it on them. There just isn’t much return for most products on the Apple OS. That’s not going to change because you don’t like MS and keep saying MS sucks over and over again.
Look, just promise me that when OSX takes the long dirt nap, you will get therapy instead of taking the bottle of sleeping pills, ok? In return I’ll remember not to click on your site’s links anymore… it’s just not worth my time.
Will Vista be the last OS Microsoft ships?
Good question. We know Microsoft has had one heck a time getting Vista out the door and that in doing so has been forced to drop lots of the original features. But it is quite possible Microsoft has learned something along the way.
Microsoft was forced to retreat to a stable baseline and start Vista over from scratch. We also know that Microsoft has begun utilizing modern-day software engineering techniques to build the new code. This is important because Microsoft’s program design has lived in the dark ages for the last decade.
One question, which can only be answered by the top Windows engineers, is did they retreat far enough. Does their base code represent a ‘black box’ that is stable but no one at Microsoft really understands or is that base code fully transparent to the coders? It makes a difference to Windows future. One problem that has plagued Microsoft throughout its Windows history is that code that no-one really understands kept getting dumped in. Each promise of a full-rewrite was abandoned as time ran out. If that core contains code that isn’t fully 100% understood Microsoft’s ability to move forward on the new code will be compromised.
Will Vista be the last OS Microsoft ships? Almost certainly not, but it is likely to be the last Windows (as we know it) OS that Microsoft ever ships. I’ve read too much from current and former Microsoft employees to make me think the Windows code is sustainable. I posit two possibilities:
1) Microsoft truly starts over and writes a new OS. This will be a tremendous undertaking and will mean that the company will probably be stuck supporting Vista for 5-6 years. Will it - can it - be backwards compatible? Maybe but don’t count on it. One reason Microsoft is in its current fix is the company’s insistence on backwards compatibility.
2) Microsoft starts with a known stable kernel and builds a new Windows compliant OS atop that. Sounds familiar doesn’t it? And since Microsoft is now researching *nix inhouse, it makes for intriguing speculation.
“Dude, ever heard of 7.5, 7.6? And OS 8 is actually 7.7… and it’s not called Mac OS 7 its System 7… it’s hard to take you seriously if you are so ill-informed…”
My point is that it took Apple 7 years to make a very large update to their OS, and Microsoft is taking the same amount of time as well. And you’re right- System 7.5 and Mac OS 7.6 (that’s what it was called in the last update) were probably two of the biggest improvements to 7, but then again, Microsoft also released two service packs to XP, so it’s not like it hasn’t been updated at all since its release.
...as for Mac OS 8, Apple changed it to “Mac OS” so they could slip out of the cloning contract they were tied into, because the contract stated that the clones could ship their computers with OSs that have “System” in the name; had you watched the 1997 Macworld Keynote, you would have known this, but I guess you’re just ill-informed, aren’t you?
Look, I’m not saying Windows is a great operating system, nor am I saying Microsoft is not at fault for their delays- I would love it if the world used OS X, but it’s just not going to happen, no matter what kind of data you back it up with, and I don’t understand why people think it will.
Hi davidwb,
I think the place where poiuytrewq goes wrong is in thinking that because Microsoft is riding high today, that it always will be. He misunderstands the nature of Boot Camp, too. Boot Camp is not a capitulation but is a flank attack on Microsoft through Dell and HP.
We are still speculating on how Apple will implement that attack. Boot Camp alone is not enough to be convincing, but what if Leopard 10.5 can run Windows-XP programs, too? That, I think, changes everything. Businesses can run proven and familiar Windows-XP programs without the vulnerabilities built into Windows. They will get all the advantages that Vista promises, but without Vista’s drawbacks. Vista will be rough for quite a while; new software always is.
I know that Microsoft is fighting back and I wouldn’t want to count them out. But, many of their past decisions have put them behind the eight ball. Companies often leap-frog each other. First and second place companies change places, because business models run out and need to be changed. New technologies are developed which must be exploited or your market share vanishes when your customers move to your competition. You must, also, position yourself to take advantage of changes that won’t happen for years and, in the meantime, you suffer for doing that.
How well is Microsoft competing now? Not well at all; it has been resting on its laurels from thinking that it’s customers are locked in to it. The Mac OS X operating system is superior to anything that Microsoft will make in the next five years. MS, to make anything better than the Mac OS, will necessarily go through the same kind of development hell that Mac OS X went through. Will Microsoft’s customers hang on during the rough times ahead? I think not. Not when the Mac offers the flexibility to run all the programs that they are used to at the same or better speeds of x86 machines, at a price and quality equal to or better than a Dell.
Another factor is the possibility of “Yellow Box for Windows.” that is, an Apple development system that can be compiled into Macintosh or Windows API’s, perhaps in a FatBinary on the same DVD. This would bypass the old “Windows has the developers” argument. Why? Because it would be cheaper to write programs for both systems concurrently and let the development system do the heavy lifting. How far away is Xcode from being able to do that? Not very far.
A lot of things are coming to a head now. Technologies are being developed that put old software standards at risk. Global competition is heating up; American companies do not have any natural advantages. The far east will be challenging those American companies.
It will take some time, but Microsoft’s near monopoly will be overturned. It is not just that Microsoft’s competitors are getting better and more numerous, but that Microsoft has made decisions that will come back to haunt them. It’s going to be interesting times ahead.
Microsoft has only delayed the inevitable. The Windows API, once the crown jewels of Windows, has now become Microsoft’s “Achilles Heel.”
Vista is still based on Windows’ current API: Win32. While Win32 was to be included in Vista, it was for backwards compatibility only. Instead, Vista was to be developed using Microsoft’s next generation API WinFX.
Microsoft successfully used the same strategy in the early 90’s when migrating between Win16 and Win32 in Windows NT. With the release of Windows NT, Microsoft could bundle multiple APIs in a single OS. NT’s core ran on Win32, but older applications could run under Win16 for backwards compatibility. Using the same technique, Vista was to be developed using WinFX, but include Win32 for backwards compatibility.
However, WinFX is an entirely new API based on .NET, managed code and C#. During development, Microsoft realized that WinFX wasn’t mature enough to host Vista and that managed code required much faster hardware to run at the same speed as XP. Instead of using WinFX, all of the new features in Vista are based on extensions to the Win32 API. Interestingly enough, Microsoft came to this realization at about the same time as development of Vista was restarted in 2003. Except for a few snap-ins and minor bundled apps, almost none of the applications shipped with Vista, including IE 7, are written using WinFX.
So what’s the problem? Microsoft can just keep extending the Win32 API, right? Wrong.
Have you ever been to Venice Italy? Instead of rebuilding, home owners in Venice can avoid construction taxes by adding new rooms to an existing house. The results may be beautiful to look at, but from structural perspective it’s complete nightmare. Win32 has been around, in one form or another, since the early 80’s. To avoid breaking compatibility from one release to another, Microsoft makes as few changes to the existing API as possible and adds new functionality as extensions. And they’ve done it again with Vista. After 25 years and 5 major releases of Windows, Microsoft has a serious mess on it’s hands.
The result? Adding features to Win32 is difficult. Very difficult. Microsoft had to cut features like WinFS and new features, such as User Account Protection, are bolted on and poorly implemented (or as Paul Thurrott calls UAP “A Joke.") Vista is XP with a shiny new GUI, updated kernel, much needed security fixes, improved sound and search, with half of the features blatantly copied from OS X.
Apple, on the other hand, successfully went though this transition in 2001. Instead of writing an entire API from scratch, Apple bought NeXT and it’s very mature Object Oriented API, renamed Cocoa. And, unlike Windows, Apple is actually Cocoa in Mac OS X. In fact, the browser I’m using right now is implemented in Cocoa and Objective-C.
While Apple may have fallen behind Windows in the 90’s, Microsoft’s failure to deliver a next generation API may have put Apple nearly a decade ahead. Microsoft will be either forced to create a completely new OS from scratch or become insignificant by continuing to release incremental updates to Windows.
Mac offers the flexibility to run all the programs that they are used to at the same or better speeds of x86 machines, at a price and quality equal to or better than a Dell
---
YOU ARE JOKING RIGHT?
Dell: £449 inc vat + del
Pentium Dual Core, 512 meg ram, 160gb hd, dvdrw combo drive, 19” TFT, keyboard mouse ect, xp home
Apple: £449 inv vat + del
intel core solo, 512mb, 60bg hd. (no keyboard, mouse, or tft)
Plus, if i want to play games on my dell machine, i can pay £200 for a decent graphics card and slot it in. Can i do this with the mac mini. NO!
You get a hell of a lot more for your money with dell im affraid.
ian, Apple doesn’t make junk that you will be forced to throw away within a few years, nor does it make low end machines. Often, PC assemblers make stripped, loss leader, low quality computers that, when you add in the features that come as standard equipment on a Mac, cost more than the Mac. Then too, there is more to a computer that it’s up-front cost. The total cost of ownership over the life of a computer cost a PC owner three times as much as a comparable Mac. That does not include the hassles of owning a PC, assuming that your time and worry is worth anything.
You could say that you only need the features that come as standard equipment on a PC. And that Dell puts together a system that fits your criteria, such as being able to plug in a decent graphics card. And that Dell doesn’t design its computers merely to conform to a check-list whipped up to induce you to buy. We disagree. But, we have different markets here. Apple does not cater to tinkerers. It caters to people who need to get work done and can’t afford the hours of maintaining a machine. The point is that, from our perspective of over-all cost, that a Dell costs more. You are free to disagree.
May I remind you that Aopen just offered for sale a knock-off of the Mac Mini; it sells for a hundred dollars more than a Mini. Imagine that.
[I’d also like to point out that the Core Solo will be a much more powerful processor than the cheapo intel dual core Ps of S, whose prices are being slashed due to their inferiority.]
I think we’re stumbling here over the fact that you’re not actually saying that they’ll just LEAVE the operating system market. This article seems slightly misleadingly titled since what you are in fact arguing is that actually that windows needs replacing with something better, or at least drastically rethinking, else it’ll be out-moded by the gradually changing face of computing.
As long as we’re using computers we’re going to be needing some kind of operating system and I don’t see the basic fact that there’s no viable alternative changing too soon.
Ben,
Windows will hang around because it has a very long tail. We talk about Microsoft Windows running on 90% of the computers in the world, but only a portion of that is in Windows XP-- between two to three years of sales. There are a few computers out there, even today, running Windows 95. The hardware wasn’t good enough to upgrade, the software, now unsupported, still works fine and the need for the system isn’t critical or used very often, so that it isn’t worth replacing yet.
Apple doesn’t want that part of the market anymore than the low end. It has it’s own long tail going back to a few Mac SE/30’s running Mac os 9 although Apple as been more ruthless than Microsoft about declaring systems obsolete. MS hangs onto the legacy equipment, partly, because once someone leaves Windows they are unlikely to go back. MS’s strategy is basically one of replacement.
John Martellaro has a very interesting article as follows.
http://www.macobserver.com/columns/hiddendimensions/2006/20060424.shtml
John is more cautious in his speculations than I; he thinks the ramifications of Boot Camp will play out over the next four to five years. I’d say sooner. His reason is that Apple R&D;has been lean and mean; that is, that Apple doesn’t have the personnel to go head to head with Microsoft. He’s right; Apple was lean and mean. But Apple has been laying on staff since the iPod became a success. It has been buying up real estate like crazy recently-- a 110 thousand sq. ft. office building near Cupertino, Offices to staff 700 customer service reps in Bangalore, India, 112 thousand sq. ft. server center in New Jersey, etc.
At the Cupertino City Council the other day, Steve Jobs was asking for approval for a new 50 acre Campus about a mile away from its present headquarters to house 30 to 35 hundred employees in about four years. Steve said that the city of Cupertino has run out of space to house the some two thousand people that Apple has added.
Regarding Scott’s excellent post, Microsoft seems worse off than I expected. I expected them to be further along than that. Microsoft needs for the world to let it catch up, but Apple is on a roll. MS has two choices: go through development hell to make a modern OS or patch what it has in Win32 to hang onto market share. I’m betting on the latter.
But, Apple won’t cooperate with that. Mac OS X keeps getting better-- it is a moving target-- ever harder for MS to catch up with. Then, there are the hardware changes coming up. and increasing competition worldwide.
I don’t see Microsoft going out of business, but that it will be squeezed from both ends. Competition with the Linux desktop will strip away the low end while Apple chips way at the top. If MS keeps fifty percent of the market, MS will still be profitable. But If Apple captures the top 25% of the market, with its profit margins, it might have a higher market cap than MS.
I don’t see many practical differences between John Martellaro and Cringley’s speculations. Running Windows XP programs inside Mac OS X should have a better user experience than running them in Vista.
John’s speculation is more of a stealth strategy; that is, it is less likely to alarm Microsoft since they are getting extra money from Apple users. Since Steve Jobs likes stealth moves and dramatic announcements, John’s speculation seems the most reasonable. Especially since, Apple often likes to put the pieces into place and then announce great changes in an offhand manner.