The Snow Leopard Report: The Law of Diminishing Expectations
June 3rd, 2009I suppose it would be nice to see your applications launch faster and your applications actually begin to use the two, four or eight cores of processor power that today’s Macs offer. That may be more than sufficient reason to buy an operating system that, for most practical purposes, otherwise offers few new features.
When I recently wrote an article about whether you even need a more powerful Mac these days, I didn’t spend a whole lot of time on the advantages. While you can’t retrieve your email any faster, nor will your Web pages come sooner, there are decided advantages to having more CPU horsepower.
Consider such basic functions as ripping a CD in iTunes. For this common task, it is absolutely true that a Mac with four cores can do up to twice as fast as the dual-core variety. What this means, of course, is that Apple has optimized iTunes to support multiple processors. It’s too bad that so many apps don’t, and that’s very true even for business software.
In the course of my work, for example, I regularly convert a pair of two-hour radio shows each week from QuickTime MP4 audio to MP3. That way most anyone can play the shows on pretty much any decent PC or music player on the planet. This is, of course, deliberate. High compatibility means more listeners, and, well, you get the picture.
None of the apps I use to make that conversion, however, seem to exercise more than one processor core on my Mac Pro, which has two quad-cores. Some run more efficiently, however. It takes Amadeus Pro, a shareware audio editor, a little over two minutes to perform the task, whereas Bias Peak Pro 6, a mainstay in the recording industry, requires as much as eight minutes. Go figure.
The arrival of Snow Leopard will mean vastly improved support for multiprocessors and it will even make it possible to offload tasks to your Mac’s graphics chip when it’s not otherwise occupied with gaming or other chores.
A you might expect, it won’t be an entirely automatic process. If an application calls on a system-related function, it will benefit from those potentially vast performance improvements. For other tasks, the developer will have to do the heavy lifting and compile an upgrade. I presume Apple will tell us that it will be a trivial process, one as trivial as (I suppose) making universal applications to run on the PowerPC and Intel.
Except that correctly optimizing and debugging an application can still take weeks or months to accomplish, so don’t expect a spate of multiprocessor-savvy software right at the starting gate. But remember that Mac OS X will be a faster beast, so there will be some improvement anyway.
Now I realize that thousands of Mac developers are already pounding away at Snow Leopard, and they could tell us many fascinating tales. On the other hand, Apple’s betas are protected by confidentiality agreements, and I would not be so bold as to ask anyone to break that contract. If you want to be an Apple developer and get the software seeds, you play by the rules and that’s the end of the discussion.
Of course, that hasn’t stopped me from reading the ubiquitous articles online that purport to reveal details about the status of Snow Leopard and, in some cases, screen shots.
At next week’s WWDC, Apple will present developers with the so-called “final beta,” and that could mean that there will be significant changes that will send them all scrambling to get their products compatible before Mac OS 10.6’s final release. Or maybe not. With Apple you can’t predict, and I won’t even try. Besides, we’ll all know the truth in just a few days.
The question mark about Snow Leopard is that, regardless of whether it stays at the standard $129 level or a lower “special price,” will a large number of Mac users be willing to get an upgrade that delivers few visible changes? Sure, the promise of much greater performance, even if benchmarked, may not be sufficient to sway most customers.
Also bear in mind that, according to the rumors mentioned above, Snow Leopard will only support Macs with Intel processors. I suppose that makes sense, if those stories are true, since it’ll come over three and a half years after the first Macs with “Intel Inside” appeared. While I suppose those of you with older Macs might feel disappointed, a lot of what Snow Leopard does requires multiprocessing and a decent graphics chip, and loads of Macs with the PowerPC simply don’t fit into that category. That’s how it is, folks.
At the same time, Snow Leopard will no doubt be preloaded onto new Macs within hours after its official release, so within a year or so the early adopter equation won’t be so significant.
One intriguing aspect of all this is how the inevitable Snow Leopard versus Windows 7 war will fare this fall. Remember, Windows 7 is largely a baked over version of Vista, offering better performance and some interface tweaks, such as a Dock-like taskbar cribbed from Apple. But you could also say that Snow Leopard is just a warmed over version of Leopard, right?
Print This Post
You have a lot to learn about Snow Leopard if you think it is a “warmed over” version of Leopard. I suggest you do some reading. Also a vast number of Mac applications are already designed for multi-processor since there have been multi-processor Macs for a long time before dual core.
If only you were correct, but if you check the Macworld SpeedMark test, you’ll see that the Macs with extra cores don’t score in multiples of the ones with lesser cores. They are faster, but not to a large extent. That’s because a still large number of popular apps don’t take advantage of multithreading and multiprocessing. That’s a big reason why Apple is doing more work in that direction in Snow Leopard.
Warmed over? Yes, and I think Apple agrees with me. While there are many under-the-hood changes, it is nowhere near as filled with new features as previous versions of Mac OS X.
Peace,
Gene
Most non geeks will not even be aware of 10.6 the same goes for Win 7. Being a true geek I will be happy to put my $129 down for what I expect to be an even more refined GUI, an OS that takes advantage of my newer hardware and to reward Apple and it’s developers for building and advancing the best operating system out there.
Gene,
Rather than speak of diminishing expectations, I think it would be more appropriate to celebrate the fact that Apple took the high road and decided to take a break from the ‘features war’ and revamp their OS under the hood. In fact Apple said right up front, that this release will not have any/many new features to market to users, so they set the expectations for that right about zero from the start.
Not only will the OS be leaner and more optimized, but according to Apple they will build the infrastructure to support new additions over the next decade or so rather than relying on their current OSX code base which dates back to 2001 or so.
The ability to make much better use of multiple cores/processors (including graphics) is also worth more than a passing mention. New computing power is growing horizontally (more cores) and not so much vertically (higher speeds).
Put me in the geek category. I’ll take under the hood changes over “features” I’ll never use any day.
It may shock people, but I’ve never used Time Machine. At first it was because the feature was buggy, then because it doesn’t work with my AirPort Extreme. It will never work for all my data because I keep my iPhoto library on a volume with permissions disabled (recommended by Apple for situations where multiple users need write access to the same library). It’s simply impossible for a single volume to back up a volume with permissions and one without. In the future I plan to use Time Machine for my boot volume and continue to use other products to back up my pictures, videos and music.
Having said that some of Apple’s high profile features over the years are terrific. I particularly like Exposé and Spaces to keep my work organized.
Snow Leopard is quickly becoming a power users OS.
While Windows users guffaw about a Taskbar that no longer limits them and superfluous window
management Mac users will be leveraging modern and open technologies and the power of Unix.
There is rapidly become a divide in computer users. The basic user, the intermediate and the expert. Today everything you read from the Tech press is aimed at the basic user. They encapsulate their articles around just the high level features that can be explained successfully to a 10 year old. Windows 7 is borderline Basic/Intermediate.
Mac users will have to utilize a few brain cells trend towards Intermediate/Expert to understand and leverage Snow Leopard technology.
The problem with reality is that it tends to disappoint people with active imaginations.
Windows 7 and Snow Leopard are not really at odds with each other because they serve different computer markets. Windows 7 is at odds with the majority of the IT personnel who think that XP is good enough. And that their present computer is fast enough.
I don’t know if you read Daniel Eran Dilger, who is sometimes hard for me to take. I disagree with Daniel’s politics and his propensity to use his column to make attacks on Conservatives, but he tends to have a good grasp on what he knows something about — technology.
Daniel says that Vista was a poor job of responding to Mac OSX, and that most of the Wintel community saw no reason to buy into Microsoft’s goals. If Microsoft’s customer base actually valued what Apple provides, why buy a shoddy copy from Microsoft?
Vista has improved and Windows 7 should be better and faster, even though it isn’t a real operating system. It is a standalone system which can’t take any punch which gets past its periphery. This is why any vulnerability hurts Microsoft Windows so much while the Mac just shrugs it off.
http://rixstep.com/2/20090601,00.shtml
Windows 7 doesn’t solve Microsoft’s problems with its customers, though. They saw little reason for an advanced compositing graphics engine such an Aqua and Areo. Or the expensive hardware which is necessary to run them. This graphics engine didn’t serve IT’s business needs, so they stayed with Windows XP.
http://www.roughlydrafted.com/2009/05/09/why-windows-7-is-microsofts-next-zune/#more-3521
What is Snow Leopard competing against, if not Windows 7? NeXT’s Openstep. When Apple bought NeXT and Steve Jobs returned to Apple’s management, Steve tried to get the Mac developers to adopt Openstep with a Mac-like cover (Rhapsody).
Steve failed to persuade them, so Apple had to create the Carbon API’s which allowed the developers to do the minimal work necessary to move to a modern real operating system which was modular, multiuser and UNIX. That job is done. The move to Intel processors has helped. Legacy hardware and software have been left behind.
Snow Leopard will obsolete the Carbon API’s or put Cocoa wrappers around them. It may take five years to get rid of Carbon or move it to an emulator, but that step is inevitable. 92 percent of the Apple user base is now on Intel, so most of the developers use Xcode, Cocoa API’s, run time languages which produce Cocoa Apps and the Objective C programing language. Those who don’t use Apple’s system run the risk of losing their market share.
Apple is doing the work of bringing Next’s Openstep up to date with the latest technical thinking. Most of this work is already there in Mac OSX, but many design compromises made in 1997 can now be discarded.
What this does is to open up new possibilities. Sun’s Zettabyte File System will allow for an infinite expansion in disk drives which will be necessary as drives expand on a distributed home computer or a SOHO wireless Local Area Network.
Apple is doing the ground work necessary for when technology turns everyone’s computer from a stand alone devise into a distributed computing system. The computer-on-a-chip will become cheap enough, so it can be put into every peripheral. But, it must operate transparently and wirelessly with the rest of the local network. This distributed computer isn’t far away now — several years, perhaps. The Netbook laptop computers are just its leading edge. Apple will not make one, because something much better than a netbook is coming.
Sixty-four bit processing does much more than merely provide a wider address space; it allows the Mac OSX 64 bit kernel to utilize the increased registers in the Intel Core 2 processors. OpenCL does more than allow the 16 cores of the NVEDIA GPU to be utilized for computations. Grand Central does more than merely provide a better utilization of multiple cores and threads. We will have dozens of computers in our homes. Trying to keep track of it all will be hard, as will deciding which device should do the work.
Many technologies which have been waiting in the wings, like VR, can start to be increasingly utilized. But, Apple will be talking about none of this until it has more of its ducks in a row. And we won’t see much of a speed improvement until more applications are recompiled in 64 bit.
Mac OSX 10.7, in 12 to 18 months, will be where the new features based on Cocoa will strike us. Apple is the leading edge. Most of Apple’s deficiencies have been resolved. Great things are in the offing, but we don’t know where Apple will be taking us yet.
Not putting the cart before the horse. Snow Leopard represents the new foundation for the next versions of Mac OS X. The processor transition began in early 2006 with a dual-core, 32-bit Intel Core. Later in 2006 was the next move to 64-bit processors. Snow Leopard’s ideal machine profile is a 64-bit, multi-core Intel CPU with a good graphics processor. The installed base of such ideal Mac has been growing since late 2006.
Windows 7 is really Windows Vista Service Pack 3. Snow Leopard for the ideal Mac represents more of a game changer with so many under-the-hood changes. Was surprised to see an October date for Win7, I thought it would have been out by now. I still see late September be the coming out for Snow Leopard and I like to believe with reduced pricing.
Win 7 hasn’t even made he switch to UNIX. Apple could cancel Snow Leopard and still be more than a decade ahead of Windows.
Looking forward to snow this summer.
DaveD wrote:
I just don’t get some people, SN represents a big improvement but I don’t want to pay $129 for it? What is up with the cheapskates on this website?
Billy
It’s not necessarily us. I’m a computer enthusiast and so I’d pay $129 for Snow Leopard no problems but the reality is we’re all in this together and just because computing is a hobby of mine that i’m willing to layout cash for doesn’t mean my Mac neighbor is the same.
The problem arises when there are more people that “don’t” want to upgrade. It puts developers in the position where they don’t want to strand people on older OS and thus they cannot fully take advantage of all the newest features.
I want Snow Leopard to have a fast uptake so that we can get apps redesigned for concurrency and better memory handling and more efficient Quicktime playback but that’s not going to happen if Apple does not deliver enough “wow” features to get the less enthusiastic Mac users over to Snow Leopard quickly.
One of the reasons for the failure of Vista was the need for something better than integrated graphics at a time when PC makers were doing everything they could to lower production costs. Tower PC buyers have the option of spending $100 on a video card and getting a night and day type improvement, but as customers move to notebooks that option is lost and they’re left with a user experience that leaves a lot to be desired.
Apple’s more efficient OS graphics layer means that even entry level Mac buyers can have a good user experience, but when it comes to taking advantage of Snow Leopard, the chips used by Apple are woefully inadequate. The Radeon HD2600 found in last year’s iMac and Mac Pro probably won’t offer any processing whatsoever and the 9400M and GT120, despite being officially supported, have such a low stream processor count that what they offer in terms of general purpose processing will be minimal.
The reality is that the ideal Snow Leopard machine, as proposed by DaveD, is in the hands of only a tiny fraction of Mac owners. Most of us will have to buy a new Mac and new applications to truly benefit.
Billy Ofstring said:
“I just don’t get some people, SN represents a big improvement but I don’t want to pay $129 for it? What is up with the cheapskates on this website?”
If it’s not worth it to them they can wait for Mac OSX 10.7 in 12 to 18 months. They can pay full price then. No skin off my nose.
I’ll get it even if Apple doesn’t have a bundle with iLife and IWorks. It should be faster and some of the 64 bit Apps should be interesting. We Mac Users tend to upgrade to the newest version if we can. Mac OSX 10.5 was introduced in August 2007 and 92% of the Mac users have upgraded to it.
What will you get with Snow Leopard? In one sense, not a heck of a lot, certainly not enough to justify the price. And by the price I don’t mean the measly $29 we now know it will cost, I mean the real price, which is the hassle of dealing with broken applications and similar problems that inevitably accompany a major system upgrade (in the past it has sometimes taken me weeks or even months to get everything back to normal). In terms of wizzy new innovations on the order of Time Machine, Spaces, or Expose, not very much at all. In fact, all I’m really goint to get is a new computer. What I mean is that with at least the three last Mac’s I’ve owned, a G-4, a G-5, and now my first generation Mac Pro, I’ve been very aware that I’ve not been able to fully tap the potential power of these machines. I’ve at least nominally had such technologies as multiprocessing and 64-bit processing, but I’ve been able to take only limited advantage of them, and the only way to get more real-world speed and power that I could actually use has been to upgrade to a new Mac. As I read the tea leaves, running Snow Leopard is going to allow me to do a much better job of tapping the latent power of my present Mac, and this is going to give me something in the general range of the performance boost I’d get if I bought a new Mac. Oh sure, this won’t happen immediately , but it will over time, as I progessively replace my current software with Grand Central-aware upgrades as they become available. If I’m even halfway right in my prediction, this is going to be the best twenty-nine buck investment I’ll ever make.
What you get with Snow leppard, dfs, is a new engine, drive train and suspension to a car that you love. It looks the same but it is peppier and drives better, although it may have picked up some eccentricities. This allows you to keep the car and improve the interior.
We will be getting 64 bit applications built in Cocoa, but perhaps, this upgrade is not for everyone. By all means, they should stay with leopard and wait untill all the bugs are worked out. They shouldn’t move until their friends kid them that they are old fogies for staying with Leopard and tell them that they are missing out on all the improvements. They probably didn’t put any stress on Leopard anyway, so why should they move?
No one will force anyone to upgrade, so why shouldn’t they delay until 10.6.4 or 5? That way they get to say that the world is going to hell in a handbasket. LOL
The point with Snow leopard is that it opens up possibilities for the future. Mac OSX 10.7, in 12 to 18 months, will be where the action will be. By that time, most people will have migrated to 10.6, the Cocoa API’s will be stable and secure. And most applications will have been recompiled for 64 bit.
That is when the power which was in NeXT Openstep can be unleased. All the compromises which Steve Jobs made back in 1997 to put a Mac-like exterior on Openstep can be discarded. Carbon API’s will be disused or put inside Cocoa wrappers. 32 bit applications will be obsoleted.
The Mac will be cutting edge, like NeXT was in 1994.
Snow Leopard is a steal.
Just look at the performance improvements listed.
Many 2x improvements in many areas.
It truly is like giving your computer a mhz bump