Daniel Eran Dilger
Random header image... Refresh for more!

WWDC 2008: Predictions & What to Expect: Mac OS X 10.6

WWDC 2008
Daniel Eran Dilger
WWDC is just around the corner, and rumors are flying about what might be revealed or released. Here’s a look at how reasonable the rumors about Mac OS X 10.6 are, along with some ideas for where Apple is headed in its desktop OS.


Mac OS X 10.6 Snow Leopard?
ArsTechnica and its sister site TUAW recently reported that Apple would unveil Mac OS 10.6 under the name Snow Leopard at WWDC, targeted at a retail release of January 2009. Rather than being a feature release, Ars’ Jacqui Cheng wrote that “it will not contain major OS changes” but instead “is heavily focused on performance and nailing down speed and stability.”

Ars also suggested that Snow Leopard would both mark the end of the Carbon API (“things that are currently only Carbon accessible will be no longer”) and drop support for PowerPC Macs.

Some of the details attributed to 10.6 make sense. For example, it seems like the perfect time for Apple to release a major housekeeping update rather than push 10.6 out further in order to add a lot of new apps and other entirely new features. At a time when Microsoft is struggling to polish the turd that is Windows Vista and is being forced to continue selling its now ancient Windows XP, Apple is coasting along with a comfortable lead in Mac OS X 10.5.3 Leopard, which it appears Apple will now call “OS X Leopard,” according to the latest WWDC banners. But take a look first at how competitive pressure can result in bad software.

Apple’s Mac OS X 10.6 code named “Snow Leopard” – report
WWDC 2008: Moscone West Spy Shots!

Getting Sloppy in Software Competition.
Software competitors are often forced by the market to release code prematurely and fix it afterward. Microsoft pioneered this aggressive “promise a lot, deliver it not” model on the desktop PC, creating an atmosphere in the 90s that said software shouldn’t really be expected to work well or even as promised. That wasn’t an accepted reality outside of the DOS PC world.

In the 80s, Apple’s Mac System 6 reached a rock solid, appliance-like stability, despite its lack of modern UNIX-like features such as memory protection. As long as you weren’t patching the OS with crazy extensions, Macs ran like serious business machines. Only in the 90s, faced with the desperation to ship new software to match Microsoft’s announcements, did Apple begin dumping out software versions that were fat, sluggish, and often unstable out of the box.

Sun OS and Solaris, along with other UNIX workstation versions, were perhaps ugly but worked like reliable tanks. Steve Jobs’ late 80s NeXTSTEP software, despite being pioneering in a groundbreaking way, was so stable and well designed that it was sought out by securities traders, the NSA and CIA, early online merchants, and others who needed advanced development tools and high reliability in an era a half decade prior to Windows 95 and NT 4.0.

Years later, both of Microsoft’s operating systems were rushed to market half finished and subsequently fatally crashed at regular intervals. Neither could maintain uptime of a month despite the vast resources Microsoft had been throwing at them. As Steve Jobs quipped last year at the Apple shareholder meeting, “I wish developing great products was as easy as writing a check. If that were the case, then Microsoft would have great software.”

Microsoft’s desktop software only got worse in the latter half of the 90s, as it pushed out more half-finished followups in Windows 98/98SE/ME, along with desperately competitive versions of Internet Explorer and Outlook email, both full of gaping security holes. Those products were all atrocious because Microsoft was sloppy, and it was being sloppy because the company was in such a hurry to suffocate any and all competitors with its stinky wet blanket so it could dominate the market. Windows Enthusiasts were happy to compromise on quality because they thought cheap hardware somehow made up for shoddy software that never worked as promised.

By 2000, once any credible competition from DR-DOS, Netscape, Sun Java, UNIX workstations, and the Mac all appeared to be out of the picture, Microsoft had the luxury of working on its code without much distraction. The result was Windows NT 5.0, branded Windows 2000. That OS was the pinnacle of Microsoft’s OS efforts: quite reliable, broadly compatible, and strongly utilitarian. Windows 2000 made it hard to complain about Microsoft; the company was finally delivering a functional product, even if it had to destroy the prospects for arguably better competitors for a decade in order to do so.

 Rd Rdm.Tech.Q2.07 60A1C88F-B504-4Cd7-Acc4-4104C9887A5A Files Applemap-3

Answers from Steve Jobs at Apple’s Shareholder Meeting
Windows 5x More Expensive than Mac OS X

The New X: an OS Battle.
Shortly afterward, Microsoft again ran into competitive pressure from early but promising builds of Mac OS X, Apple’s reincarnation of NEXTSTEP. Microsoft scrambled to nip Apple’s advances in the bud by branding Windows 5.1 as XP, suggesting a relationship in the level of sophistication between Apple’s OS X and Windows XP. However, Microsoft’s XP simply layered all sorts of consumer fluff from Windows ME on top of Windows 2000, resulting in a heavy, cartoonish new version of Windows derided as a Fisher Price user interface. It also compromised the strengths of Windows 2000 to add broad compatibility with consumer devices and games, as well as the insecure Active X and related initiatives intended to weld the web to Windows.

In contrast, Mac OS X started with the enterprise-proven foundation of NEXTSTEP, and added legacy support for existing Mac OS applications by cleaning up the old Mac development APIs under the name Carbon. It also incorporated significant new technology including Quartz, the first sophisticated graphics compositing engine on a desktop PC.

Since XP’s release in 2001, Microsoft has been forced to revisit security and stability in Windows XP and attempt to shoehorn in .NET as a modern new development API to replace the archaic, labyrinthine mess of Win32. While Microsoft kept busy doing this, Apple had the freedom to rapidly develop upon its core foundation with new reference releases of Mac OS X in Jaguar, Panther, Tiger, and Leopard, each of which pushed compelling, marketable features while extending performance and enhancing its developer tools and APIs.

Microsoft’s belated response in Windows Vista (6.0) was intended to catch XP up with Apple in terms of desktop sizzle, but the project was such a huge undertaking that it was destined to fail. It also suffered from Mac OS X envy. Apple’s users are computing mavericks who go out of their way to choose something new; Microsoft presides over a sleepy, utilitarian user base that buys Windows largely because it is the easy default option that appears to be cheaper. Most Windows users don’t want redesigned flash, they want a functional, basic system that works well enough at the lowest price possible.

And so, as anyone with any background in software development should have foreseen, Windows Vista has been a year long failure since its release while Apple’s Tiger and Leopard releases of Mac OS X have captured fawning media attention. Microsoft is now left telling its consumer base that it will catch up to the iPhone and Leopard and Apple’s other advances with Windows 7, a release that can’t credibly happen before 2010. It too is fated to disappoint everyone, as the minority of Windows users who want bragging rights to coolness will never see Microsoft eclipse the smaller, faster Apple, while the mass of corporate users who just want utilitarian functionally will be irritated by Microsoft’s efforts to chase Apple’s stylish taillights.

 Wp-Content Uploads 2008 04 200804221422

Windows Vista, 7, and Singularity: The New Copland, Gershwin, Taligent
Windows 95 and Vista: Why 2007 Won’t Be Like 1995

Rapid Advancement vs. Building Strength.
Microsoft labors with both the concepts of fixing its insecure and sloppy legacy and in keeping up with the fashionable sizzle of Apple’s designs. Meanwhile, Apple has advanced both its interface and its core technology. Mac OS X 10.2 Jaguar, 10.3 Panther, 10.4 Tiger, and 10.5 Leopard all advanced new flashy features, but each also delivered major new performance strength underneath.

As everyone from body builders to Risk board game players knows, you can only make rapid surface advancements so far before you have to stop and rebuild your core strengths. Each major reference release of Mac OS X has done that. The idea that Mac OS X 10.6 will focus more on strength than flashy new advancement is interesting, but it’s highly unlikely it will happen in terms of how it is being presented by the Ars rumors. Apple does have the luxury to deliver 10.6 without needing to push lots of new applications, but it also has clear plans to deliver major new OS features.

Had Apple pursued a strategy of advancing fluff features over the last two releases, it would certainly be time for some core OS strength training. But that’s not the case. The flash in Tiger and Leopard was not layered on top of a weak foundation, as Vista’s glossy UI was over XP. Instead, all of the new features advanced in Mac OS X have been implementations of core OS strengths Apple has incrementally built.

The translucency and shadows in Mac OS X are not just eye candy, but a way to demonstrate the power of its underlying graphics compositing technology, which has led Windows’ conventional bitmap graphics for seven years now. Prior to Vista, XP apps could only fake translucency and other powerful compositing effects Mac users take for granted. By the time Vista delivered Microsoft’s implementation of Apple’s Quartz engine, its new smoky glass effects were really nothing to crow about. Instead, the new graphics processing overhead (that Mac OS X has been running all this time and XP hasn’t) just made Vista look slower on the same hardware.

Similarly, Apple isn’t finding itself in the same Win32/.NET conundrum as Microsoft. Vista promised to deliver all its bundled apps as managed code built on .NET, from its Finder-like shell to its other included apps. Microsoft found it couldn’t quite pull that all off, leaving Vista to just run updated versions of the old Win32 XP apps, with many of the old holes from the 90s preserved intact, such as the ancient printing security bug that hit Vista right as it left the gate.

Apple cleaned up its original Mac OS API in Carbon over a half decade ago, and developers have been getting advancements to it and Cocoa ever since. Apple has increasingly strengthened Cocoa rather than just promising to use it. QuickTime 7 completely overhauled the video architecture to dump old legacy related to QuickDraw and add support for the latest technology in video compression with H.264. Apple also delivered Cocoa QT Kit tools that allow developers to make advanced use of its powerful audio and video features, and uses them itself in its own software. Consumers just saw everything work faster and better.

death of yellow box

Cocoa and the Death of Yellow Box and Rhapsody
WWDC 2007: Kevin Hoffman Presents .Net vs. Cocoa

Thinking Inside the Box.
Anyone who thinks that Mac OS X 10.6 will be an enormous bug fix that just makes Leopard faster and less likely to present any errors is writing with refrigerator magnets. Apple can’t advance along its same trajectory without adding key new features and significantly advancing major core OS changes. One core change already in progress relates to kernel managed security features.

The iPhone’s managed, signed code will make it highly resilient to efforts to inject malware to run on it. That restriction is enforced in the OS X kernel. Apple has made it clear that the desktop version of Mac OS X will get similar treatment. That likely won’t be fully in place in 10.6, but more of the foundation will be. Apple recently filed for patents in that regard. Microsoft is working in similar directions, but has more developers to get on board. Apple has some cats to corral too, but its smaller, tighter platform will enable it to get developers to play along easier.

Apple already writes much of the software many new Mac users run: Safari, Mail, iLife, iWork, and Apple’s other bundled apps will pioneer the concept of signed, manage code, allowing for secure parental controls, managed policy in corporate circles, and general “sandboxing” limitations enforcing what downloaded apps or scripts from untrusted sources can do without being explicitly assigned permissions. This is a significant new direction, already in progress, that will be fleshed out in 10.6. Everyone is working on managed code, but there are also other unique directions Apple is taking in 10.6. The next segment examines Future UI Designs in Mac OS X 10.6.

iphone signing

iPhone 2.0 SDK: How Signing Certificates Work
iPhone 2.0 SDK: Readers Write on Certificate Signing

WWDC 2008: New in Mac OS X Snow Leopard
I really like to hear from readers. Comment in the Forum or email me with your ideas.

Like reading RoughlyDrafted? Share articles with your friends, link from your blog, and subscribe to my podcast! Submit to Reddit or Slashdot, or consider making a small donation supporting this site. Thanks!

Technorati Tags: , , , , , ,

  • Pingback: WWDC 2008: Moscone West Spy Shots! — RoughlyDrafted Magazine()

  • Pingback: WWDC 2008: Future UI Designs in Mac OS X 10.6 — RoughlyDrafted Magazine()

  • Pingback: Apple News and Information » Blog Archive » Alert - apple rumor release()

  • gus2000

    “[Microsoft] was in such a hurry to suffocate any and all competitors with its stinky wet blanket so it could dominate the market.”

    Nice imagery. “Windows SWB” lol

  • Dalekrium

    I think Ars would be right after a fashion.
    Apple would release in this WWDC a pre-release 10.6 that would not have any DEVELOPER new functionality.
    Remember this is a Developer Conference and with each version of OSX new Functionality for the programer has been presented CoreAudio, CoreImage, CoreVideo, CoreData, CoreAnimation, QTKit, KNI Api.
    I think in 10.6 no new Core functionality will be presented and the 10.6 pre-release will not present user upgrades of importance asking the developers to center its work with this alpha in stability an security. And maybe a extension of Cocoa to cover the corner functionality where Carbon is still needed.
    I go as far as saying that the pre-release distributed at WWDC will be intel-only.

    With the actual plan been that:

    In January the pre-release with the user upgrades added will be seeded and presented in MacWorld and in May-April the Gold master of 10.6 will be reached.

  • lmasanti

    I think that there will be some backward feed from some of the developments that arose from the iPhone.
    As for a release date, I vote for March 24th… Mac OS X anniversary.

  • roz

    I really don’t think XP was a response to OSX. I think that was Msft’s branding – they thought it was cool at the time on its own.

  • Berend Schotanus

    Dan,

    I already expected you to announce this one: “One core change already in progress relates to kernel managed security features.” and here you go!

    Yes this is interesting, absolutely fascinating, because it can change the business model for computer software. I would expect Apple is progressing stepwise and very careful. The introduction of iPhone apps might have been prepared long before when Steve Jobs wants it to look like. Now they will try out how it really works on the iPhone, how software developers react, how customers react. They will adapt, fine tune the model. And when it works well, why not restrict Mac software to signed software. And what about the role of Apple’s retail channel, the iTunes store, in distributing this software …?!

    Apple has to move slow because there are a lot of interests at stake in software business.

  • Pingback: 2008 release of iphone()

  • elppa

    I think it was. Windows up to that point had been named after the year. I thought what became XP was going to be Whistler.

    Although they will never ever admit it, I think Dan’s “me-too” naming theory might not be a million miles away from the truth.

  • Pingback: under eye circles()

  • http://johnsessays.blogspot.com John Muir

    Thanks for not nodding in unison along with all the other pundits and bloggers on this Snow Leopard rumour.

    It sounds like several oldies rehashed and served up as a nice WWDC-eve stew.

    Ingredients:

    1. Leopard is the new Vista. Ergo: “Leopard is broken and needs a service pack!”
    2. Adobe must die. Therefore: “Adobe must too!”
    3. Leopard’s delay was humiliating. So: “Apple are rushing out the next thing to save their blushes!”
    4. Windows 7 > 10.6. Hence: “10.7 by 2010. Hell, even 2009!”

    Here’s how I see it:

    1. A sneak preview of 10.6 *may* be given at WWDC.
    2. The next OS X will be more comprehensively Cocoa, including the Finder and Apple’s Pro apps.
    3. Carbon IS NOT being taken out back and shot. If it is, so too are Photoshop, Illustrator and Office. Apple has given it a timeline of doom, but won’t kill Carbon apps any release too soon. The Cocoa thing is about *their* software, not Adobe’s.
    4. More code will be shared between iPhone and Mac, at every level. Synergy between the platforms will be emphasised. Indeed, they will ultimately become one.

    Alas, such details as Apple’s software strategy are lost on far too many Mac commentators, yet alone TUAW. Written in refrigerator magnets indeed … wakey wakey Jacqui.

  • http://johnsessays.blogspot.com John Muir

    Addendum: ingredient 2 was Carbon. Adobe just so happen to be its most substantial devotee.

  • lmasanti

    quote:
    “And when it works well, why not restrict Mac software to signed software. And what about the role of Apple’s retail channel, the iTunes store, in distributing this software …?”

    The new store is called AppStore… not iPhoneAppStore.
    So, why they should use iTunes store to distribute software if they have a whole store to sell software?

  • http://coderad.net StrictNon-Conformist

    Just one thing I’d like to add in comparison to Apple and the Mac OS X mutations compared to Microsoft and Windows mutations:

    The biggest strength of the Windows OS for Microsoft is the overwhelming (though not perfect) backwards compatibility. This allows people to keep using the software for long periods of time, unmodified. This provides a lot of stability that a lot of people demand of their infrastructure.

    The biggest weakness of the Windows OS is also the same thing: this makes it nearly impossible for them to move forward quickly and make any meaningful changes to how things work, if it would break software compatibility. The amount of testing that needs to be done to verify that is immense, even for Microsoft and all their people.

    The biggest strength of the Mac OS X variants (and this was also true of the pre OS X Mac OS) is that Apple can more readily break away from backwards compatibility: this allows them to make changes faster, and keeps things comparatively lean by not retaining backwards compatibility hacks.

    The biggest weakness in the corporate world of the Mac OS has also been its biggest strength: corporations don’t like replacing their software frequently, so this makes it more of a chore to keep things updated. Thus, you have people working with stuff like Adobe that’s now having problems because the platform is moving forward without it coming along as fast as it should. Adobe software isn’t custom in-house software: it’s more general market, and has a much larger budget available to update it to the latest API mutations, and if they’re being responsible, they’ll do it proactively. However, constant modification for new mutations of an OS and its API just isn’t something that’s easy to justify to bean counters for in-house software.

    So, you have two choices: pay the price of bloated backwards-compatible Windows OS mutations that makes hardware groan, as well as developers, as a result of the byzantine API, or keep on the treadmill of OS X for keeping up to date, and not having long-term backwards compatibility. They both have their prices, and the question is whether you want to pay one price or the other, or you can afford to.

  • worker201

    @StrictNon-Conformist:

    What you say is true. However, I think that at some point, concessions have to stop being made. There are companies out there who still use DOS-based software, and are trying to run it on Vista! I don’t know which is more foolish – trusting your business to 25-year-old technology, or updating past Windows 98 if you want to use DOS. It is folly to expect full OS support for such extreme cases. In many ways, I think Apple is right to pressure its users to move along and leave outmoded crap behind. Of course now that I’ve said that, my favorite apps will not be 10.6 compatible ;)

  • Pingback: WWDC 2008: Is Mac OS X 10.6 the Death of PowerPC? — RoughlyDrafted Magazine()