Apple

Apple and Microsoft are fighting for the mindshare of consumers as both companies prepare to roll out upgrades to their operating systems later this year.

Apple on Monday showed Worldwide Developers Conference attendees Snow Leopard, the next major version of Mac OS X. Apple has been very open about the fact that Snow Leopard is meant to be an under-the-hood maintenance release, focusing on performance enhancements to the operating system.

Windows 7 is essentially Microsoft's maintenance release for Vista, that according to many accounts was a failure for the company. Putting aside all of the back and forth between the two companies, one industry analyst feels it comes down to the consumer.

"It's really immaterial the degree of the rewrite in the operating system," Ross Rubin, director of analysis for market research firm NPD, told CNET. "The key is the consumer benefit."

Apple introduces Mac OS X Snow Leopard at the WWDC.

(Credit: Jim Dalrymple)

While early testing of Windows 7 seems to bear out improvements in the operating system, Microsoft is coming off a very bad consumer experience with Windows Vista. That is not a trivial obstacle for it to overcome.

Microsoft will have to fight the industry perception that Windows 7 is just Vista with a few fixes. That could certainly lead to slower adoption of the new operating system out of the gate.

Apple on the other hand is coming off one of the most successful operating system launches in the company's history. Mac OS X Leopard was a solid release, packed with features. Overall, Leopard had relatively few problems throughout its life cycle.

Apple doesn't have to fight off that negative perception from its users or the industry. Macs have been selling better than ever and there is no sense that will slow down anytime soon.

Typically, Apple sells its new operating systems for $129. That's a flat fee. Everyone gets the same version that includes all features and enhancements. However, Leopard users will be offered an upgrade to Snow Leopard for $29. Microsoft has yet to release its upgrade pricing, but it is expected to be much higher.

"The OS war is on in a big way," said Michael Gartenberg, vice president of Strategy and Analysis at Interpret. "Charging $29 won't win Apple any converts, but Microsoft is going to look really bad with its upgrade pricing."

It's clear that Microsoft has a much bigger channel to push Windows 7 to customers, but we've seen with the Vista release that doesn't always mean success for an operating system.

Apple is coming from a strong position with Mac OS X Leopard, so upgrades to its newest Snow Leopard release should be very strong.

Apple said Snow Leopard is expected to ship in September. Microsoft will release Windows 7 in October.

Apple has fired back in the latest round of the Mac versus PC ad wars with two commercials tweaking Microsoft's marketing strategies.

One nice thing about having a sick girlfriend is a guilt-free weekend in front of a dozen or so college and pro football games. Apple released two new Mac vs. PC ads for that sedentary audience to ponder in between kickoffs this weekend, though both ads seemed to be tailored more for the tech industry than NFL fans.

Both commercials poke fun at Microsoft's recent massive ad campaign to "redefine" itself in the face of two years of clever Mac vs. PC ads that have helped Apple increase sales of the Mac. One commercial has John Hodgman in his now-familiar role as PC allocating stacks of bills toward either "advertising" or "fix Vista." Guess which pile gets more stacks.

The other commercial is also about Vista: Hodgman has developed a buzzer that bleeps out "Vista" whenever that word is uttered, so that people will start using the term "Windows" instead of Vista. This is actually funny, given that Microsoft was recently awarded a patent for similar technology, even if that wasn't the point of the ad.

Apple is trying to do two things with the ads: get under Microsoft CEO Steve Ballmer's skin, and continue to define Vista as a glitch-ridden operating system. Vista's early problems with application and driver compatibility are well-documented, and while most of those problems are in the rear-view mirror, Microsoft is spending more time these days talking about stereotypes, Windows 7 and the "Apple tax" than it is about Vista.

Apple, on the other hand, is quite content to keep Vista in the firing line.

The fact is that the negative impressions of Vista have stuck, as Microsoft itself had to admit with the Mojave Experiment. Whether that's Apple's fault or your IT department's fault isn't really the point; Apple took the early reluctance of people to try out Vista and ran with it, while Microsoft sat on its hands for two years then vacillated between saying "Vista isn't so bad," "we've been unfairly stereotyped," and "yeah, you might want to wait for Windows 7."

Still, negative advertising, while effective in an election year, grows old. And it plays into the sorely outdated fanboy us vs. them mentality that the vast majority of consumers couldn't care less about; most people in America do not define themselves by the computer that they use, as hard as Apple and Microsoft are trying to make that happen.

The new ads will get a chuckle out of most viewers, as the polished comedy team of Hodgman and Justin Long could teach Seinfeld and Gates a thing or two. But God forbid that Apple should ever stumble with the rollout of a new operating system; they've taught Microsoft just how to respond.

Apple made a few jabs at Microsoft during Tuesday's notebook event, but if I were a Windows executive today, I'd probably be breathing a sigh of relief.

Although Apple did revamp its entire Mac product line on Tuesday, it didn't hit the $800 price point that the rumor mills had projected.

The new laptops seem nice enough, and they might be enough to keep Apple on a roll, but Apple didn't take the Mac into any new segments of the market.

Clearly fearing that Apple was on the brink of such a price move, Microsoft launched a pre-emptive attack on Monday, with Vice President Brad Brooks going into great detail about an "Apple tax" ahead of what even he thought would be the introduction of significantly cheaper Macs.

It seemed at the outset of today's event as if that might be where Apple was headed. In his introductory comments, Apple executive Tim Cook rattled off several reasons Macs are doing well. One of those things on the list, he said, is something Apple has nothing to do with--Windows Vista.

But in the end, Apple's attack was limited to those words and the company's usual arsenal of elegant but pricey machines. Apple has made some significant advances, to be sure, but it looks as if this year's Mac-vs.-PC battle will remain at the high end of the market rather than dropping down to the mainstream.

Nonetheless, it is interesting to watch the war of words heat up. For years now, Vista has been a popular punching bag for Apple, and Tuesday's event was no exception.

Microsoft, though, is only belatedly trying to defend itself, with the most visible effort being its "I'm a PC" advertisements.

The reasons for Microsoft's moves are clear. First of all, it had let itself be completely defined by a competitor. Secondly, Apple's market share is significant and growing. Although its share of the global market is only a few percentage points, its share of the dollars spent on PCs, particularly in the United States, is far more significant.

For complete coverage of the Apple notebook news, see "Apple polishes up its MacBook line."

SAN FRANCISCO--Politicians like to joke that Social Security reform is considered the "third rail" of politics. In Apple's world, that rail belongs to security.

It's been a while since we examined the "state of Mac security," and with this week's RSA Conference in San Francisco, and last month's CanSecWest conference fresh in everyone's mind, it seemed like a decent time.

The topic is always a heated one, and it tends to bring out the usual Mac vs. PC bashing. But according to people I talked to this week here at RSA, the nature of security threats has moved well beyond the platform.

First of all, let's examine where things stand. No security researcher I spoke with could think of an instance of a Mac running Mac OS X that had been exploited in the wild. Not as part of a contest, or as part of a show-stopping demonstration, but through a malicious attack aimed at pwning a Mac. Few were even sure that any viruses or worms existed for the Mac; there was a Trojan horse type of exploit in the wild last year, but it was delivered through a porn site, and it required users to take several steps to infect themselves.

So Macs remain a very safe computing option. This does not mean that Mac OS X is secure, however. It's software, written by humans, and it contains flaws. Those flaws are theoretically exploitable by criminals, but they haven't been, mainly because you don't need an MBA to do a cost-benefit analysis.

Apple hasn't had its "come to Jesus" moment yet with security, the way Microsoft did in the early part of this decade. Millions of Windows users demanded that Microsoft fix the leaky boats that were Windows XP and Internet Explorer, and to Microsoft's credit, it stopped almost everything it was working on and set about that task.

That hasn't happened to Apple. Even though Apple's market share continues to grow quarter by quarter, the company's products account for just 5.8 percent of the total U.S. market for PCs, according to IDC.

Charlie Miller pwns a MacBook Air at CanSecWest last month.

(Credit: TippingPoint)

"Market share equals money" to the hacker criminals of the world, according to Charlie Miller, a researcher at Independent Security Evaluators. Miller made headlines last month by taking control of a MacBook Air as part of the CanSecWest conference's "Pwn to Own" contest. He used a previously unadvertised flaw in Apple's Safari browser to gain control of a system that was directed to a malicious Web site, earning himself and his team $10,000 and a new MacBook Air.

"Even if Apple moved to 10 percent market share, why spend the time on the 10 percent when you can just nail 90 percent with one bug?" Miller points out. It's far easier, and far more lucrative, for those shadowy figures in the hacking business to spend their time going after the other 90-plus percent of computers in the world than it is to try to exploit flaws in the Mac--even if there's a shiny new computer involved.

Changing of the threat
More and more, it's not really about taking control of a computer through flaws in the operating system; it's about using the browser as the entry point into the system or hacking Web sites, said Mike Romo, product manager for Symantec's Mac product line. "Trojan horses and viruses are yesterday's news."

At the CanSecWest conference, no one was able to take control of three laptops in play (the MacBook Air, a Fujitsu running Windows Vista Ultimate, and a Sony Vaio running Ubuntu) when attacks were confined just to the operating system. But Miller's Safari exploit, and the Flash flaw later exploited by Shane Macaulay, Derek Callaway, and Alexander Sotirov on the Vista laptop, show how security threats are now much more focused on the browser, rather than the operating system.

And it's also much more about phishing and social engineering your way into someone's wallet than it is about trying to take over their system, Romo said. "The OS is not really the target anymore for these next generations of threats; it's taking advantage of the fact that people are spending more time online. People are much more comfortable with entering a credit card number than they ever have before," he said.

That means it's no longer about Windows vs. Mac OS (at least when it comes to security debates--don't worry, fanboys). It's about Internet Explorer vs. Firefox vs. Safari vs. Opera. It's also about things like QuickTime, which Apple has patched extensively since the "Month of Apple Bugs" project last year.

Symantec distributed some research this week showing that 22 vulnerabilities were reported for Safari in 2007, compared with 88 in Mozilla browsers like Firefox, 18 in Internet Explorer, and 12 in Opera. It should be noted that counting the vulnerabilities is not the best way to measure the security of a piece of software, and can be explained in part by increased interest on the part of security researchers in investigating Firefox and Safari, as they become more widely used.

Browser flaws, not operating system flaws, are increasingly the more dangerous entry point.

(Credit: Symantec)

And, as Symantec points out, "as security researchers have focused more efforts in discovering vulnerabilities in these browsers, the theory that this would result in much greater levels of malicious activity targeting these browsers in the wild has not yet been borne out." Again, IE is still the leading browser, and it makes more sense financially to go after that product.

The problem for the security industry is that even if Microsoft, Mozilla, Apple, and Opera all make the most secure browser ever, it still won't prevent things like phishing scams. The quickest, and perhaps easiest, way to make money from criminal activity on the Internet these days is to send out one of those Nigerian 419 e-mails, have people visit a Web site and enter their information, and shut that site down after a few hours of gaining credit card numbers.

It's almost impossible for security companies like Symantec to track that kind of quick behavior and update browser protection software to recognize the phisher's site as a threat, before at least a few people are affected. Lather, rinse, repeat, and after a while, you'll take in far more cash for a day or so of work than you would toiling away for weeks trying to exploit a flaw in Vista or Mac OS X, Romo said.

This is as much a social problem as a technical one; lots of people who may already be nervous around computers often just do whatever the computer tells them to do, Romo said. Credit that tendency for some of the uproar around Apple's decision to ship a new version of Safari to Windows users through Software Update. More than a few people didn't realize that they didn't have to do what the computer was telling them to do.

Miller and Romo--both Mac users--worry that the need for greater security to protect people from themselves will force Apple to change the way the Mac handles certain tasks, potentially taking away some of the Mac's ease of use. Leopard already takes a step in this direction, Miller noted, though not nearly as far as the User Account Control feature introduced in Vista, to much derision.

But Apple's not going to adopt Microsoft's security strategies for Mac OS X, until users demand it or hackers force its hand. They simply don't have to. Until then, quick, diligent patching and a wider embrace of the security community will more than do its part in keeping the Mac secure.

Education and "safe surfing" practices are as important to this era of security as anything having to do with counting flaws or patching practices. Maybe that's the third rail of technology writing: it's not always the mean evil corporation's fault; sometimes, it's yours.

It held out as long as possible, but a Windows Vista laptop fell to a determined bunch of hackers Friday evening at the Pwn to Own contest at CanSecWest.

Since it was the third day of the contest, which saw a MacBook Air get hacked on Thursday, the TippingPoint Zero Day Initiative relaxed the rules even further. On the first day of the contest, only the operating system could be targeted, but on the second day that was expanded to include standard applications. An undisclosed Safari flaw led to the MacBook Air's downfall.

TippingPoint's Aaron Portnoy, with Shane Macauley and Alexander Sotirov (left to right) take control of a Windows Vista laptop.

(Credit: TippingPoint)

But on Friday, hackers could target any "popular" piece of application software that you might find on a system. The Fujitsu laptop, running Vista Ultimate, was compromised by a previously undiscovered flaw in Adobe's Flash software.

Shane Macaulay, Derek Callaway and Alexander Sotirov, were able to gain control of the laptop, which also means they get to keep it. However, since the rules had been relaxed, they only get $5,000; the MacBook Air winners collected $10,000.

The contest rules stipulated that any winner sign a nondisclosure agreement immediately after a successful hack, so that the nature of the flaw could be disclosed to the vendor. Once Adobe and Apple patch their flaws, the nature of the flaw will be disclosed.

A Sony Vaio laptop running Ubuntu remained unscathed at the end of the conference.

We updated this blog at 6:25 p.m. PST after Microsoft released a statement.

As far back as 2005, Microsoft executives knew that confusing hardware requirements for the Windows Vista Capable program might get them in trouble. But they did it anyway--over the objection of PC makers--at the behest of Intel, according to e-mails released as part of a class-action lawsuit pending against Microsoft.

In early 2006, Intel's Renee James, vice president and general manager of Intel's software and solutions group, was able to prevail on Microsoft's Will Poole to change the proposed requirements for Microsoft's proposed "Vista Ready" marketing program to include an older integrated graphics chipset that couldn't run Vista's Aero interface. At the time, Intel was worried that it wouldn't be able to ship the more advanced 945 chispet, which was capable of running Aero, in step with Microsoft's proposed schedule for the introduction of the marketing upgrade plan.

This led to the creation of the "Vista Capable" logo, which is the reason Microsoft is now in court, facing a class-action lawsuit on the part of PC owners who bought so-called Vista Capable machines in late 2006 only to find those machines could only run Vista Basic, which doesn't feature the Aero interface. The potential for confusion was well-understood both outside the company, as noted here in this CNET News.com story from March 2006, and within the company, as multiple e-mail threads reveal.

Intel's Renee James, head of the chip maker's software and solutions group

(Credit: Intel)

A treasure trove of e-mails has been released as part of that case, and the Seattle Post-Intelligencer's Todd Bishop has spotlighted a number of e-mails that call into question whether Microsoft was acting, at least in part, on Intel's behalf when it set the requirements for the Vista Capable marketing program. (Read all the e-mails released by the court in this PDF.) Several pages of e-mails were redacted by the court. All e-mails quoted in this report were taken verbatim, typos and all, from a PDF file put together by the Seattle Post-Intelligencer in a blog posted by Bishop yesterday.

"In the end, we lowered the requirement to help Intel make their quarterly earnings so they could continue to sell motherboards with the 915 graphics embedded," Microsoft's John Kalkman wrote in a February 2007 e-mail to Scott Di Valerio, who at the time managed Microsoft's relationships with the PC companies and recently took a job with Lenovo. The change took place in January 2006, and was formally rolled out by Poole, currently corporate vice president of Microsoft's unlimited potential group, without the knowledge of Jim Allchin, the now-departed Microsoft executive who was supposed to be in charge of Vista's development

Intel declined to comment on specific e-mails until it had a chance to review them. But in response to the Kalkman e-mail, read to a Intel representative, the company said, "We do not know who John Kalkman is. We do know that he is not qualified to know anything about Intel's internal financials or forecasts related to chipsets, motherboard or any other products. He would have no visibility into our financial needs in any given quarter."

Jim Allchin, Microsoft's former head of Windows development, currently retired

(Credit: Microsoft)

The planning for the Vista Capable program started long before it was publicly announced in May 2006, a few months after the final delay in Vista's ship date was announced. The idea was to mimic what Microsoft did with Windows XP, to assure customers buying PCs sold within a few months of the launch date that their hardware could run the new operating system when it was formally released. This helps PC makers avoid a swoon in demand in the weeks and months prior to the launch of a new operating system.

Microsoft knew that Vista's Aero interface would put a significant strain on the hardware used in those PCs, and so in 2005 it started putting requirements together for the Vista Ready program using Intel's 945 chipset as the baseline chipset needed for designation as "Vista Ready."

Eric Charbonneau, an unidentified Microsoft executive, told his direct reports in August 2005 in an e-mail that the older 915 chipset wouldn't cut it. "Any OEM who plans to ship an Intel 915 chipset system (using UMA, without separate discrete graphics hardware) for Summer 2006 needs to know that: 1. Their systems will not be eligible for the Windows Vista Ready designation..." Simply put, the 915 chipset couldn't support the Windows Vista Display Driver Model (WDDM), and that capability was a requirement at the time for being able to slap a "Vista Ready" sticker on a PC.

However, at some point between that e-mail and January 2006, Microsoft changed its stance on the 915 chipset. The 945 chipset was Intel's top-of-the-line integrated graphics chipset when it was introduced in May 2005, but it still sold lots of lower-end 915 chipsets in both desktops and notebooks. Intel didn't launch the notebook version of the 945 chipset until January 2006, and was apparently concerned that it would be unable to get enough 945 systems into the market by the middle of 2006, the (at the time) launch expectation for the Vista Ready program.

With notebooks a far-faster growing segment of the PC market than desktops, Intel apparently felt that if only 945 chipsets were deemed Vista Ready, that demand for systems with 915 chipsets--still a significant mix of its products--would fall off the face of the earth. And also, that it would be unable to produce enough 945 chipsets to meet its committments to PC makers--orders that might otherwise go to Advanced Micro Devices.

In January 2006, Poole sent an e-mail to several Microsoft executives informing them that the plan had changed, and that Intel approved. "I went over the new plan with Renee tonight. Not surprisingly, she is pleased with the outcome. I told her we wanted to communicate to OEMs and retail first, and then they can cascade their own communication. They are losing orders every day, so we need to get a simple communication out ASAP."

In February 2006, one month after Will Poole informed the Vista team of the decision, Microsoft's Will Johnson wrote an e-mail laying out some more of the specifics.

"We have removed the WDDM requirement for Vista Capable machines, the modern CPU and 512 RAM requirements remain intact, but the specific component that enables the graphical elements of Windows Vista (re: aeroglass) has been removed. This was based on a huge concern raised by Intel regarding 945 chipset production supply and the fact that we wanted to get as many PCs as possible logo'd by the 4/1 US retail REV date. The push to retail should be that while this opens up a wider band of machines to being Vista Capable retailers should be very aggressive in communicating to their OEMs (and thus Intel) to maximize production of 945 chipset equipped machines going forward."

According to e-mails exchanged, many inside Microsoft were appalled at the decision to let Intel's supply concerns dictate its marketing policies. Now Microsoft had to go out and create a two-tiered program promoting both "Vista Capable" machines and "Vista Premium Ready" machines.

A Vista Capable sticker would simply mean the PC could run Vista Basic, allowing PC makers to promote their PCs as "Vista" PCs while glossing over the fact that the minimum hardware requirements for that label couldn't really handle the improved graphics that were one of the major reasons to upgrade to Vista. This confusion was exactly what Microsoft and its PC partners hoped to avoid when they were first drawing up the requirements in the first place, and several e-mails show those concerns were shared widely prior to, and following, Poole's decision.

Hewlett-Packard was particularly incensed, since it had decided to adopt Intel's 945 chipset more aggressively, believing it was the only chipset that would support the Vista Ready program.

Microsoft's Mark Croft wrote in response to Poole's e-mail that, "We need good messaging for the elimination of WDDM in Capable, as we have had this as a requirement since inception over 18 months ago."

But perhaps the most surprised executive inside Microsoft at the move was Allchin, the head of the Vista development team.

"We really botched this," he wrote in a thread responding to Poole's e-mail. "I was not involved in the decision making process and I will support it because I trust you thinking behind the logic. BUT, you have to do a better job with customers that what was shown here. This was especially true because you put me out on a limb making a commitment. This is not ok."

Will Poole, co-head of Microsoft's emerging markets efforts, who authored the e-mail acknowledging pressure from Intel.

(Credit: Microsoft)

Later, in a private e-mail, Mike Ybarra of Microsoft pleaded with Alchin to step in and reverse the decision. "Jim, I am passionate about this and believe this decision is a mistake," he wrote. "We are caving to Intel. We worked hard the last 18 months to drive the UI experience and we are giving this up."

Allchin appeared to agree in his response, but seemed resigned to fate.

"It might be a mistake. I wasn't involved and it is hard for me to step in now and reverse everything again," he wrote to Ybarra. "We might be able to thread the needle here if we make 'capable' just related to 'old' type hardware."

And so, confusion began, just as Microsoft employees and partners predicted it would. Some Microsoft marketing units started saying that the even older 865 chipset would now qualify for the Vista logo program, which was squashed. But it was easy to see where the confusion stemmed once the requirement for WDDM was dropped, as essentially anything relatively modern that could easily run Windows XP would be capable of running Vista Basic.

Anantha Kancheria wrote to Rajesh Srinivasan as part of a discussion in March 2006 around the 865 confusion, and employed a little gallows humor.

"Based on the objective criteria that exist today for capable even a piece of junk would qualify. :) So based on that yes 865 would qualify. For the sake of Vista customers, it would be a complete tragedy if we allowed it. I don't know how to help you prevent it."

The 865 was eventually scrubbed from the program, but the 915 was allowed to remain. And so, PCs with the 915 chipset were sold as Windows Vista Capable, while others sold with the 945 chipset or better were labeled Vista Premium Ready. As predicted, confusion ensued, and even Microsoft executives and directors were snared.

Steve Sinofsky, the former head of Microsoft office development and current head of Windows and Windows Live development, wrote an e-mail to Microsoft's Brad Goldberg in July 2006 asking about a Dell Latitude he purchased that he thought was labeled as "Vista Ready," but in reality didn't have enough graphics hardware to run Vista.

Steven Sinofsky, Microsoft senior vice president, Windows and Windows Live Engineering Group

(Credit: Microsoft)

Goldberg, then vice president of Windows product management, explained, "Some PCs that are windows vista capable will run aero and some will not. In the interim we've created a marketing designation that allows OEMs to market PCs as "premium ready." every pc that is premium ready will run aero."

Goldberg continued, "for holiday oems will be heavily pushing premium ready machines but because Intel was late with their integrated chipset the majority of the machines on the market today are windows vista capable but not premium ready. originally we wanted to set the capable bar around aero but there are a bunch of reasons why we had to back off...a bit messy and a long story that I'm happy to walk you through if helpful. :)" Goldberg has since been reassigned.

In January 2007, Jon Shirley, a former Microsoft COO and current member of the board of directors, wrote CEO Steve Ballmer an e-mail complaining about driver support for some peripherals he wanted to use with his Vista PC. Ballmer forwarded the e-mail to Sinofsky, asking for input on whether Microsoft should be doing anything differently.

Sinofsky launched into a post-mortem on Vista itself, with this graph pertaining to Intel.

"Intel has the biggest challenge. Their "945" chipset which is the baseline Vista set "barely" works right now and is very broadly used. The "915" chipset which is not Aero capable is in a huge number of laptops and was tagged as "Vista Capable" but not Vista Premium. I don't know if this was a good call. But these function will never be great. Even the 945 set has new builds of drivers coming out consistently but hopes are on the next chipset rather than this one."

Ballmer's response? "Righto thanks."

Microsoft is now defending itself against claims the Vista Capable program was misleading and unfair, all thanks to a decision to allow Intel to sell older chipsets that couldn't run Vista's Aero interface--really one of the main reasons to upgrade--with the word "Vista" attached. As the e-mails show, many within the company knew they were heading down this path when they embarked on a two-tier logo program, but the need to keep Intel happy--over the objection of the world's largest PC maker--won out in the end.

UPDATED: 6:25 p.m., PST - Microsoft issued the following statement after this blog was posted: "We included the 915 chipset as part of the Windows Vista Capable program based on successful testing of beta versions of Windows Vista on the chipset and the broad availability of the chipset in the market. Computers equipped with this chipset were and are capable of being upgraded to Windows Vista Home Basic. Microsoft authorized the use of the Premium Ready designation on PCs that could support premium features of Windows Vista."

About 9 percent of the Mac OS X installed base upgraded to Mac OS X Leopard over the weekend, according to figures released by Apple and estimates supplied by financial analysts.

Apple sold 2 million copies of Leopard between Friday and Sunday night, which includes sales of boxed copies, online sales, and new Macs with Leopard preinstalled. When Apple launched Tiger, it took the company 39 days to hit the 2 million mark on a much smaller installed base.

Apple sold two million copies of Leopard between Friday night's launch event and Sunday.

(Credit: CNET Networks)

Piper Jaffray released a research note Tuesday estimating that the Mac OS X installed base is around 23 million users at the moment, as compared with the 12 million Mac users that existed in April 2005 when Tiger launched. At the 2007 Worldwide Developers Conference, Apple CEO Steve Jobs said the Mac OS X installed base was 22 million users.

It's tough to get a direct comparison, but because everyone loves comparisons, Microsoft sold 20 million copies of Windows Vista in the first month it went on sale. Earlier this year, Microsoft CEO Steve Ballmer said the Windows installed base would be around 1 billion users by the end of the company's 2008 fiscal year, which comes to a close next June. And as of last week, it had sold 88 million copies of Vista.

Likewise, it's tough to get a sense of the current installed base of Windows PCs. If you subtract some generous projections for PC shipments over the next nine months from the 1 billion target, you get around 800 million Windows users. That's far from a perfect number, however, since many of the PCs that will be purchased between now and next June will most likely replace older Windows PCs.

Still, people in emerging markets are buying PCs for the first time--and buying a lot of them--so the expansion of the Windows installed base does move in step with the expansion of the overall market to a certain degree. Let's split the difference and assume there are 900 million Windows users as of right now. That would be mean that almost the same percentage of Windows users have upgraded to Vista in nine months as the percentage of Mac users that upgraded to Leopard in two days.

Let's be clear: these are different types of markets, and they have different motivations for upgrading. Almost all the Vista upgrades so far have been consumer and small business purchases, and businesses will probably start to upgrade in greater numbers next year following the release of Vista Service Pack 1 and as the PCs bought during the boom years of 2004 and 2005 begin to age.

Still, it's pretty clear the Mac installed base was more excited about Leopard than the Windows installed base was about Vista. Maybe that's because Mac users hold onto their machines for a longer period of time and like to upgrade, while Windows users just wait until they need a new PC. Or perhaps it's because Apple does a better job marketing its releases, or because Windows customers have been trained to wait until the first service pack comes out before upgrading. Any way you slice it, however, it's interesting stuff.

Progress is measured in steps both big and small. The smaller ones may get less attention, but they are much easier to take.

It's been a year of big steps for Apple. The company dropped the "Computer" from its name in January as a way of showing Apple was no longer just about the Mac, and the clear priority for 2007 in Cupertino was to get the iPhone out the door and selling briskly. Then, perhaps for kicks, it decided to overhaul its entire lineup of iPods.

Later today, Apple will take a smaller step, with the launch of Mac OS X 10.5, code-named Leopard. Leopard's coming on scene later than expected, almost 30 months after Tiger (Mac OS X 10.4) launched in April 2005, in part due to the push to get the iPhone out in time. CNET'S review is in, and my colleagues Elsa Wenzel and Robert Vamosi are positive.

Leopard, the latest version of Mac OS X, goes on sale at 6 p.m. today wherever you live.

(Credit: Apple)

There are dozens of important new features in Leopard, perhaps most notably the Time Machine application that could make it easier for users to back up and restore their files. Backing up your files is generally a simple exercise with a external hard drive, but Time Machine is interesting because of the friendly way in which it lets you restore files, flying back in time (and space) to the last instance in which that file was saved.

But all the reviewers, including Apple favorites Walt Mossberg at The Wall Street Journal and David Pogue at The New York Times, felt compelled to point out that Leopard is very much an evolution of previous versions of Mac OS X, and not a dramatic breakthrough like some past releases. It's certainly nothing like the tectonic shift Microsoft users went through in the switch from Windows XP to Vista, or Windows 98 to XP.

That can come off as a negative assessment. But it's not.

Computing trends change so quickly now: are you doing the same things with your Mac today that you were when Tiger was released in 2005? Maybe, but you're definitely capable of doing much more today, and even more so compared with when the first version of Mac OS X arrived in 2001. With Leopard, Apple will have made five major upgrades to the original Mac OS X operating system in six years.

Guess what other operating system made its debut in 2001. After the launch of Windows XP that year, it took Microsoft a well-documented eternity to release Vista, during which it changed its goals for the operating system several times and wound up releasing a solid, if underwhelming product earlier this year.

Here's the lesson: making smaller, more frequent changes to your product makes it much easier to stay on top of a changing industry than a five-year plan will ever allow. It keeps engineers on their toes and also makes the bean counters happy. That's because modest upgrades can be released more frequently that still have enough new bells and whistles to justify charging for the new software. A new copy of Leopard, for example, will set you back $129.

As my colleague Ina Fried noted to me as we watched the World Series on Wednesday night, Microsoft does make incremental changes to Windows. But it calls them service packs, and it gives them away for free. Apple sits in a happy middle, where it can make substantial--yet relatively modest--additions to Mac OS X, charge more than $100 each time, and have customers walk away satisfied that the upgrade was worth their time and money.

Of course, life is different for Microsoft. As they add new features, they have to make sure everything plays nicely with a 20-year history of code, so their business customers don't freak out. This makes it much harder for Redmond to turn on a dime to respond to new trends like mobility or multimedia.

Leopard's a run-scoring double, to stretch the baseball analogy. It's not a revolution in Mac software, but it's a nice advance for older Mac owners as well as those new to the company in the last year or so. If Apple can get back on the 12- to 18-month pace of new releases that CEO Steve Jobs told The Times he'd like to stick to over the next several years, Apple could be able to pick out the next trend in personal computing well ahead of Microsoft if the engineers in Redmond stay on their current schedule.

The age-old Apple-Microsoft debate is changing. Microsoft continues to run a very profitable business, and even if Apple continues to expand its Mac market share, Windows will remain by far the dominant PC operating system when this decade ends.

But Apple has more momentum, as the iPhone and iPod continue to make both the mobile phone and music industries take notice. It has more investment, now worth more than tech-industry stalwarts IBM and Intel but still well behind Microsoft. And it's more nimble, a crucial advantage as an era dominated by the PC comes to a close and something new takes its place.

I'm not touching the Vista versus Leopard question until I've had a chance to use them both more thoroughly. But is Leopard a more significant advance compared to Tiger than Vista is compared to Windows XP? Nope. And Jobs is probably fine with that.

SANTA CLARA, Calif.--If you're not exactly sure what you want in a mobile computer, don't worry: the folks who are building them aren't entirely sure themselves.

The consensus among five panelists gathered here at the ARM Developers Conference was that this is a very interesting and confusing time to be thinking about the future of mobile computing, because the playing field is so wide open and because consumers haven't decided exactly what they want.

"It's sort of like Darwin," said Tony Milbourn, director of mobile devices at Motorola. "We don't know what people want, we put them out there and see what people will buy."

This is about the quest for the next big mobile computer, something more attractive than a UMPC but more powerful than a Treo. It's been a very common topic of late, with the craziness attached to anything related to Apple's iPhone and Intel's clear goal of throwing its hat into the mobile computer race.

The iPhone is very much on everyone's mind (at ARM's press conference earlier in the day, executives from about six different companies had a picture of the iPhone in their presentations), but more will be needed if regular people are going to embrace true handheld mobile computing.

There's three technologies that must evolve for this to happen. Motorola's Milbourn thinks that bandwidth speeds have to improve to allow mobile applications to flourish. Other panelists, such as Jorgen Behrens of Symbian and John Lilly of the Mozilla Foundation thought it was all about applications and the user interface. And obviously, the hardware is going to have to deliver sufficient performance at battery-friendly power levels.

"The operating system is very important, but it's mostly important for the people making the devices," said Behrens, executive vice president of marketing for Symbian. The combination of the browser and the user interface dictate whether or not people will enjoy their experience, he said.

That suited Lilly just fine. The chief operating officer of the Firefox development organization thinks that the browsing experience is going to be extremely important for mobile computers, especially as people rely more and more on Web-based applications, like Facebook, Google and countless others. The problem is that right now, the memory footprint needed for an advanced browser to support those Web applications is way too large. Mozilla is working on a solution to that problem, and this reliance on Web applications could make the debate over third-party applications on the iPhone moot, he said.

Web applications also bypass the problem of operating system fragmentation in this world, according to several panelists. One reason (among others) that Microsoft came to dominate the market for PC operating system was the need to have a common platform for applications in a non-networked world, Milbourn said. But this industry is evolving in a very different way.

"There's an extraordinary awareness of not handing Microsoft the keys to another kingdom," said Jim Ready, CEO of MontaVista, which earlier in the day signed a collaboration deal with five other ARM licensees to work on Linux products for this category. "(The fragmentation) may indirectly benefit Microsoft if you think you need a real common platform with a lot of applications. But if you're on the Internet, the local platform isn't as common."

Despite Intel and Microsoft's interest in future mobile computers, don't expect the scenario that played out more than 20 years ago to happen again. "This is not going to be the PC market," said Mike Muller, CTO of ARM. "There is going to be diversity and I don't think there's going to be one product or one winner."