Post a Comment
Nothing too exciting for average home users.
I can understand their energy to start focusing more on the business users, but unless Snow Leopard has some convincing features home users won't bother to upgrade. Although this could be their strategy anyway, to skip a user friendly release and focus more on business and engineering user types.
I'm still happily running on OS X Tiger and have no intention on upgrading, until I buy a Intel mac in the next 18 months.
I guess this means everything will be Cocoa-based.
Really, this seems more like a sort of "10.5 Service Pack 2" than a totally new release. They'll charge money for it because they know that the Mac fanatics who care will actually pay for it. Heck, if it's not too expensive (which IMHO it shouldn't be considering the lack of new features) I might consider upgrading, too.
As for whether it's Intel-only, I'd say there's quite a good chance. By the time it comes out, almost three years will have passed since the first 64-bit Intel Macs were put on the market. Snow Leopard could be Apple's way of keeping the customers with older Macs happy since they're not missing out on much with "normal" Leopard, while gently pushing towards the day when those machines are obsolete--probably about two years after Snow Leopard comes out.
I was thinking the same thing -- HFS is kind of a big joke as far as file systems go (normalizing unicode characters and all sort of other fun treats)
Mac OS X + ZFS would be pretty sweet. "The server edition will have ZFS, according to Apple.
http://www.apple.com/server/macosx/snowleopard/
"ZFS: For business-critical server deployments, Snow Leopard Server adds read and write support for the high-performance, 128-bit ZFS file system, which includes advanced features such as storage pooling, data redundancy, automatic error correction, dynamic volume expansion, and snapshots."
If Microsoft spent a whole development cycle just cutting bloat, would you buy it? I suspect many would. Removing the bloat is what people have been asking for for years, and the *day-to-day* improvements would be the reason it would be worth it.
Leopard isn't exactly bloated, as it still manages to be a little faster than Tiger; but if your computer could operate twice as fast, just with a software upgrade - how much are you willing to deny that that would make a refreshing difference on the *day-to-day*? Less hard disk space? Faster graphics via the GPU? Better load balancing across 8+ cores (especially when your time=money).
Some have labelled this as a "maintenance" release.
Spending the full engineering capacity of 1000's of developers for a whole development cycle, is not "maintenance". It's full-blooded development for the future.
Apple want the OS that matters most to regular users - "is it fast?". And we geeks can say we want features, but if anything it still boils down to the same thing, we use the computer to it's full, we want it to be fast.
I really think this will be the most interesting and exciting OS release for geeks by far. It may not seem it yet, but by the time release is near, it's going to become a real talking point.
Agreed.
And more, with the "decline" of the Windows, comparison of Leopard with Vista will surely be the done concerning speed/bloat/design (as it's already done).
Maybe MS realized that already with Windows 7 being a more polished Vista ?
Even though geeks want features they/we represent a small fraction of the whole computer selling... Average joe wants things they/we(?) can use...
Leopard isn't exactly bloated, as it still manages to be a little faster than Tiger; but if your computer could operate twice as fast, just with a software upgrade - how much are you willing to deny that that would make a refreshing difference on the *day-to-day*? Less hard disk space? Faster graphics via the GPU? Better load balancing across 8+ cores (especially when your time=money)
Dude, it's not going to run twice as fast. 15% would be a huge accomplishment, probably closer to 10%.
Well it's a question of what you're doing.
Likely on average the machine is going to run 0% faster because most client machine time is spent waiting on a user. But for the latency of certain actions, which is what users care about, it is indeed possible to extract huge performance gains, depending on how optimized the particular scenario is.
Also when scaling up, if you're at a cpu count where your scaling becomes poor due to lock contention, you can easily extract 50-100% gains by breaking the locks.
50%-100% would be extreme corner cases, the vast majority of the time threads just don't get used enough to make multi-cores make a difference. Dual Core does make one, as you (more or less) are using one core for what you are doing, the other for the os. Quad Core is a complete waste for 95% of uses that a home user computer is used for.
Aparently all support for single core was dumped from 2k8, and honestly I don't notice much in the way of perf difference between it and vista 64 as a workstation machine.
Not all OS follow the MS School of threading. A true SMP based OS would run any thread on any processor/core. In this scenario, the quad core looks like a good idea again, no?
Personally I think the name gives it away. It's another Leopard, it's not another cat. They may call it 10.6, but it is basically what Windows 7 will be for Vista...
Personally, I think this is good news.
The more the core OS can move away from Carbon the better.
But it will still run on 32bit machines and will run on PPC too...
However, the optimisations will most likely be specifically targeting the Intel 64 multi-core platform, so the biggest gains will be there.
I expect it will be free too, especially as it is still a Leopard... If it will be sold, then either the speed increases will be mind blowing, or they will need to add some enticement...
I'm not sure about you, but I don't have 16TB of RAM and I'm guessing most home users don't, so that's not going to be much of a draw card...
I think an OS is too big to download, and too much of a change to release on unsuspecting users through software update. If they don't charge the full price for it, then I would suggest something like a $29 upgrade price for Leopard users, to get the DVD through the post.
For me this is exactly what I want. Can any of you remember the last time one of the large OS'es focused on performance rather than new features? Give us proper java and I am sold
The OS is suppose to be there to support the applications...not steal valuable resources away from the applications (*Vista*).
Edited 2008-06-10 07:53 UTC
exactly my words... i would love to see one more thing though: some kind of optimization for SSDs, low-latency/high-speed
maybe ZFS too
You want ZFS read/write? You get ZFS read/write!
Bottom left: http://www.apple.com/server/macosx/snowleopard/
For business-critical server deployments, Snow Leopard Server adds read and write support for the high-performance, 128-bit ZFS file system, which includes advanced features such as storage pooling, data redundancy, automatic error correction, dynamic volume expansion, and snapshots.
Too bad they couldn't get rid of enough bloat to get it practical on laptops or desktops. Looks like they might require that 16TB of RAM to run it.
i would have to echo the above comment (hangloose) Microsoft is doing exactly the same thing with windows 7. They said they are optimising it and that Vista has had all of the major changes they are now working on fine tuning. And just as apple will charge for 10.6 microsoft will make you pay for windows 7.
I also would have to agree that i would pay for the upgrade, i think it's a nice breathe of fresh air that both apple and microsoft are ditching the feature race and are both working on creating faster, leaner OS's
Although i do find it quite amusing how both companies seem to be following the same track at the same time. That both leopard and vista introduced new features and both are working on optmised follow ups.
Also ive heard that exchange support is coming in 10.5.4, which is quite interesting and would make sense with the iphone v2 update
Windows Vista's biggest competitor is Windows XP (not OS X - uptake for Vista is slow largely because of people sticking to XP not switching to Macs). Seeing that by the end of the month, XP sales will stop, by time Windows 7 comes out, it would be a choice between Vista and 7.
For enterprises, which is Microsoft's biggest customer group and the sort of base Apple is trying to target with Snow Leopard and iPhone 2.0, I think Windows 7 stands more than a fighting chance.
Here's the thing: I don't think Apple meant Snow Leopard to be the answer to Windows 7, rather trying to get into the Windows XP replacement business. Snow Leopard is coming out next year - Windows 7 in 2010 the earliest. It's a good strategy I must say, as a lot of enterprises are vary of Windows Vista.
I think more and more enterprises replacing hardware will note that low-end Macs running Leopard way better than similarly spec PCs running Vista. And this proposition will become better with Snow Leopard (and with Exchange 2007 support, they don't need to migrate their mail servers too).
The period between XP's death and the release of Windows 7 is Apple's best shot at getting a comfortable foot in the enterprise market. I'd say they're making a really good, and probably successful, effort.
Have to agree with the general sentiment that it's a good thing.
OSX is beautiful on the desktop, however it's not so great underneath. The raw performance is laggardly compared to other platforms. Things like the filesystem and the kernel are not highly regarded.
I just hope as part of this effort they provide a nice stable core for third party developers. I'm keen to see a (non X11) OpenJDK port as well as more development tools and languages.
As for new features, I'm not entirely fussed. A better performing core with a better filesystem would probably lure the cash out of me.
Pity that I have a 32bit Core Duo though. That would kill my enthusiasm. I will most likely buy another mac, but probably not in that timeframe.
Has anyone more information about OpenCL? I searched the web, but it seems to be a (proprietary) Apple technology. And the now botan-called open crypto library project is surely something different.
OpenCL would be great, if it was actually open as it name suggests. I would surely love to have a nice, compatible (i.e. no Nvidia/AMD-only library) library to offload operations to the GPU.
see my posts above and
2:11: Demo of Grand Central Dispatch (thread multiplexer) - fundamental change to Unix standard, automatically managing thread channels based on "pressure"
salvaged from the google cache
and
http://daringfireball.net/linked/2008/june#tue-10-snow_leopard
well, is it an extension to c or to object-c?
Edited 2008-06-10 21:33 UTC
Apple says it will be an "open" spec, which implies not proprietary. Hence the "open" in "OpenCL". Who knows. Apple has done this with a few technologies, but not often.
Edited 2008-06-11 07:10 UTC
I think Apple are doing a performance release like this to lull Microsoft into a false sense of security. They know Windows 7 will probably get the same reception as Vista, so they're coiling up, ready to strike, with either 10.7 or .8, or maybe even Mac OS XI. Purely speculative, of course. Microsoft will see a sluggish uptake of Windows 7, and Apple will strike when the iron is hot: just as Windows 7 starts to pick up momentum. The upgrade treadmill has been slowed a little, and a decisive blow to Microsoft, such as breaking the market momentum of their flagship product after the lackluster performance of Vista, could break the treadmill once and for all. Or at least change it's ownership.
I, for one, welcome our new, artistically inclined overlords.
There's no way an OS release will take out Microsoft. Especially if it's relegated to running on a specific hardware platform only.
You want to get Microsoft where it counts? Woo the corporate user base to switching to Mac. Make a Mac that's more affordable than the Wintel counterpart for enterprise users. And have a server product that can actually compete with Active Directory.
In going for Intel, it amazed me at the time that Apple went for 32bit hardware when they could have gone for 64bit.
Some aspects of OSX would still have had to exhibit 32bit APIs but it would have been worth doing for twists in the road like 'Snow Leopard'.
The 'Snow Leopard' features being touted are none too earth shattering, so there must be more to come.
Each new OSX edition kills off older models with insufficient capability; and I wonder where the divide will fall this time. My guess is that G4's will be out, G5's will be in for one more lap; and some features will depend on your GPU (okay for MacPro, less so for MacBooks).
This G4 Powerbook probably won't make the grade so will 'end of life' when 10.7 appears. This sets it's life at about 8 years.
There's nothing to be amazed about, the only processor they could put in a notebook was an old 32 bit G4 without hope to have a 64 bit G5 ready in a reasonable amount of time, because IBM has more lucrative customers than Apple.
Intel, instead, gave them a precise plan.
This G4 Powerbook probably won't make the grade so will 'end of life' when 10.7 appears. This sets it's life at about 8 years.
Leopard killed my beloved 800 MHz iBook G4, which was less than four years old. I wouldn't be surprised if its snowy brother kills my "new" first generation MacBook, which will be less than three when it will be released.
I wonder if it'll really be a free upgrade, coz Apple tends to charge for everything these days. Here in my country, all that "similar costs compared to a WinPC" talk doesn't apply and Macs along with their software command a chunky premium.
If Apple expect me to pay for what seems to be a fancy 'service pack', just a year after I paid 10k+(SG dollars) for my Mac Pro, it had better be damn bloody fast.... anything slower than Vector Linux (blindingly responsive on a budget laptop) and I'd consider it a disappointment. I'm already up to my neck paying for new camera bodies, new printer/paper/inks, ....etc.
BTW, does Apple even sell a Mac that can hold 16TB RAM?
Given that you say Vector Linux is so blindingly responsive on a budget laptop and you appear to value "blindingly fast" so much, why not just install Vector Linux on your Mac Pro and be done with it?
Edited 2008-06-10 09:45 UTC
Tell me for my own benefit what Linux will do on a Mac Pro that people buy Mac Pros for, that outstrips what OS X is used for?
Video editing?
Highend Photoshop work?
Audio processing?
The only thing I can think of is science work, and Leopard is Unix already.
Having a super powerful computer is one thing, but buying a Mac Pro not for OS X confuses me? If you wanted Linux for power, you'd build a power machine.
That's like buying a Bugatti Veyron so you can run it on Diesel.
That's exactly my point.
If Vector Linux is so good and is Teh Snappiez and that's what's important to you, why not run it on the Mac Pro and stop complaining about Mac OS X being slow?
If you're using a Mac Pro, it's most likely because you are using Mac apps that have no equivalent on Linux. If that's the case, an OS upgrade that makes your machine run faster which in turn makes you complete tasks faster will boost your productivity. When productivity increases, $$$ flowing into your pocket increases.
Even if the upgrade costs full price at £80 (dunno what it is elsewhere), you'll more than likely get a significant return on that small investment.
evangs,
Your telling me about Productivity??
It takes me the same amount of time in Leopard as it did in Tiger to convert the RAWs, manipulate/retouch in Photoshop, import to Corel Painter and paint over it, import back to Photoshop to sharpen and touch-up, open Genuine Fractals to interpolate, load profile and print on my Epson. I also produce the same quality prints regardless of OS version.
Eyecandy means nothing to me if it interferes with my workflow. Would Snow Leopard really make me a better Photographer/Artist/Fine-Art Printer?
Leopard runs well on my current Mac Pro, but I can't help feeling it's really bloated. Using Vector (or any other Xfce distro) as an example.... programs open with the same speed (or faster) on a low-end laptop as it does with a Mac Pro that costs 6-8 times more. Of course, I can't do Photoshop benchmarks coz Adobe don't make a Linux version.... but you get my drift.
What I'm trying to say is:
Apple seem to tout the words 'Speed' and 'Stability' with this new release. I'm all for those two qualities, but paying someone to remove bloat that shouldn't have been there in the first place is bit too much in my opinion. Of course, it's better than paying someone like Microsoft to add bloat.
I'll definitely continue using Macs coz, 1. Adobe, Corel....atc don't make software that runs natively in Linux and I don't have the time to re-learn the open source equivalents like Gimp, and 2. Colour Management in Linux is a bitch, even for seasoned users.
Perhaps I've become cheap since I started using free software..
Just my 2 cents.
That's like buying a Bugatti Veyron so you can run it on Diesel.
http://www.revver.com/video/750256/audis-diesel-supercar-street-leg...
Granted, it's no Veyron, but if we're warping analogies this badly, who's to care? Seriously, it's like calling The Pro "if money were no object, this would be the most technically advanced model of its type anywhere on Earth"-esque. Nice as the Pro is, that just ain't so.
edit: spellcheck
Edited 2008-06-10 13:28 UTC
Code optimization is apple's best defence against linux
The reason I say this is that, from what ive seen, open source development its nigh on impossible to get your developers to stepback and optimize everything that has gone before. Im not saying the coding is bad in the first place, but people are far more willing to contribute their free time makeing 'superNewFeatureX' then slogging through to optimise someone elses code..
Imagine KDE 4 having no new features, only optimisations... It wouldnt get the developer support
Edited 2008-06-10 10:20 UTC
I think that's a little unfair. There must be many unsung heroes who do real grunt work making Linux run in minimal memory and function on proprietary hardware.
Just because their name isn't presented to you when you log in, it doesn't mean that there isn't people, both paid and volunteer, who are doing the tough work too.
What Apple are doing is the marketing of that process.
There will be new features in Snow Leopard (Finalised ZFS for instance), but they won't be marketed at the forefront until maybe the final hurdle.
I agree with apple using marketing to their advantage
I also agree with the fact there are linux devs who do grunt work (they must exist or my desktop wouldnt boot in under a week :-p)
But, open source tends to have a latest and greatest mentality (just like its users). Getting all the related projects which make up a booting system to stop adding features and only optimize for a 6 months would be an incredible achivement to say the least!
Its a shame because I think we are reaching the point in linux where something like that needs to take place..
You tend to hear a lot about the greatest and newest on a news site. Those who use red hat or slackware do not have this mentality, or they would use Fedora. they're just in the news or on news sites because they're using old stuff.
I don't believe linux need a rework on performance. It beats BSD on many fronts. Or are you talking about a specific distro of GNU/linux?
Just because it beats BSD on many fronts does not means there is no room for improvement ;-)
THe point I was trying to convey is because of the structure of many open source OS you couldnt universally say 'only optimisation for 6 months' and expect it to happen
Apple (and MS) have this luxury as indipendant companies
If nothing one's gotta love Apple's love for detail, down the symbolism of the name, it is the same cat just an albino, white and clean so to say.
If I would have to make make one thing thats better in Apple products, than it's not so much more features or a better product they make, it's just that after they make one, they themselves treat it like a princess.
It's the main thing I miss in Windows, just look at the 100 age old Win3.1 add Font dialog in Vista, or the Icons on the desktop still redrawing and blinking in Aero, just lacks so much love.
I think that should be a factor in a every review like
Stability:
Security:
Features:
Usability:
and Love:)
Anybody else find this text from the apple website hilarious?
"With Snow Leopard, the next major version of the world's most advanced operating system, Mac OS X changes more than its spots, it changes focus."
How different from another famous focus shift... One company dying and moving away from performance/sleekness on their core platform, another extremely healthy, moving toward more performance on their core platform! With language like that you almost wonder if it is a subtle jab...
Edited 2008-06-10 12:38 UTC
In this case, Apple aren't doing anything special. During the development of Leopard, there was supposed to be a new kernel, which is probably what they're showing here. Avie Tevanian, the V.P. of R&D, left around the same time, maybe in disagreement over the direction. As I recall, that's been well over a year.
Who knows, maybe they're dropping Mach for some other micro kernel?
So Apple is not adding in more useless apps, but spending the needed time to make the system better and more responsive?
Good Call Apple. 2 thumbs up.
Not many OS Manufacturers listen to there user base about the system slowing down or such, but you did. Bravo.
Now for all of you whining about not having new apps / etc. then just stop and think about it. a better, more responsive system will, if nothing else, increase the life of your hardware for quite a while.
name me one other developer that has even bothered to try to do that.
hell, even Linux has become bloated. I remember back when a full install of Linux w/X included, was only 500Mb.
So yeah, Apple is doing a good thing here.
I give Apple credit for doing the right thing! Taking the time to optimize and rearrange the structure of the OS so that future OS innovations may be implemented more easily. Sometimes it takes balls to do the right thing!.
Leopard is a very stable OS and IMO, not bloated like other OS's (see Vista & most popular Linux distros). Some people seem to associate optimization with bloat reduction... Whiles this can be the case, I doubt this is Apples intentions as Leopard is not terribly bloated. It is more likely going to take the existing OS and restructure certain features/functions in a more logical/efficient way.
The great thing about OS X is that I can install Leopard on my fathers 6-YO 1Ghz dual PowerMac G4 Quicksilver and have it run great! Try installing Vista on a 6-YO computer and see what happens!
MS is attempting to un-bloat win7??? I really doubt that a company that spent 5+ years putting together an OS, removing virtually all the initial "money features" in the name of "just getting it done", is actually going to be able to make large strides is optimizing the OS. I have very little comfidence in MS abilities to do anything substantial with their OS.
If Apple really wanted to improve the performance of OS X and applications... It would write its own very highly optimizing compiler that specializes in producing code designed to run on multiple processors. And then it should do all it can to encourage/force developers to write threaded applications that make good use of multiple(4+) cores.
Krreagan
Edited 2008-06-10 17:44 UTC
This is precisely why people buy Apple computers and OSs, code reviews.
It's a "Do the Right Thingâ„¢" attitude.
When possible - do it once, do it the right way.
Commercial OSs are not mission critical like what is written by NASA for the Space Shuttle:
http://portal.acm.org/citation.cfm?id=287023&dl=ACM&coll=po...
http://www.nasaspaceflight.com/content/?cid=5026
But ... it is nice to know they're thinking of us.
heise provides some information about great central:
http://www.heise.de/newsticker/WWDC-OS-X-10-6-versteht-sich-mit-Mic...
german only, sorry. great central seems to be a library and great central dispatch somethink like a thread-scheduler which can break up a programm in as many threads as needed if the programm code contains "blocks" - an new feature of objective-c.
or at least that's what i imagined from the describtion.
according to apple:
performance, ~50% faster than leopard, 20% less CPU load
http://twitter.com/juicepharma







