Snow Leopard Q1 '09
Now that it's been leaked that Snow Leopard's target date is Q1 '09 what are some of your thoughts? How can Steve market features that are mostly "under the hood"? WIll ActiveSync be a part of the Exchange support or will Apple simply create an HTML based email client like Outlook can do with Hotmail?
Comments
Now that it's been leaked that Snow Leopard's target date is Q1 '09...
Leaked?!
http://www.macrumors.com/2008/11/18/...ue-in-q1-2009/
It comes from a PDF presentation by an Apple employee.
http://www.apple.com/macosx/snowleopard/
They don't actually say since when though. If it's about a year since the Leopard release or if it's about a year since the site was posted. It's not legally binding so if it's early, all the better.
In a way delaying the desktops will be good as it means that I don't have to buy Snow Leopard separately.
I am not buying it.
Ditto...I'm putting this rumor in the same pile as the 2GHz PPC, resolution independence by the end of 2008, and ZFS for everybody pile.
With that said, Hubbard may not have had explicit clearance to say that so it may not be 100% accurate. He may know when the OS is supposed to be finished in engineering, but may not have baked in the rest of what would need to happen before shipping. Or maybe he has? Either way it isn't rumor, it is Apple provided public information.
Flat out - this is not a rumor. It is a stated fact by an Apple senior engineer, Jordan Hubbard, at an industry conference. This wasn't an partner company employee mouthing off like those PPC and Sun zfs comments, this was Apple's Director of Engineering of Unix Technologies. Hubbard is pretty senior in the Apple engineering food chain.
With that said, Hubbard may not have had explicit clearance to say that so it may not be 100% accurate. He may know when the OS is supposed to be finished in engineering, but may not have baked in the rest of what would need to happen before shipping. Or maybe he has? Either way it isn't rumor, it is Apple provided public information.
I wouldn't go that far...an Apple provided public info/statement comes in the form of press releases or a web page on www.apple.com. Anything else is just people "mouthing off". I don't care if they're senior or CEOs of said company, it's not an official source.
You completely contradict yourself. The rock-solid counter example - every Stevenote. Everyone of those is solid gold, few points are actually followed up with press releases. And those Steve messages, coming from that meaningless CEO of your description don't always come at Apple events.
Please adhere to your self-imposed forced ignoring of all public Apple comments which are not published and ignore everything Jobs, Ives, Chubby, et.al. have to say from here on out.
Nice to see you as irrational and wrong as ever.
You completely contradict yourself. The rock-solid counter example - every Stevenote. Everyone of those is solid gold, few points are actually followed up with press releases. And those Steve messages, coming from that meaningless CEO of your description don't always come at Apple events.
Please adhere to your self-imposed forced ignoring of all public Apple comments which are not published and ignore everything Jobs, Ives, Chubby, et.al. have to say from here on out.
Don't worry. I *always* ignore anything Steve Jobs has to say. But by all means, *you* should continue to lap up his every word. *lick* *lick* MMMMmmmmMMMM. Good huh?
By solid gold, did you mean solid shit? Sorry if I'm not as delusional as you are.
Two little notes in a presentation are very little to go on so it's not worth going to great lengths analyzing it but I think it would be good if Snow Leopard has reached a point in development where it could be released in Q1.
I really don't see why it couldn't be. Apple have said themselves that they are focusing on optimization. Plus look at the release schedules of the other systems:
10.0 - 6 months
10.1 - 8 months
10.2 - 9 months
10.3 - 14 months
10.4 - 19 months
10.5 - 21 months
The Tiger and Leopard releases were for two systems - Intel and PPC. The PPC-only releases took under 14 months. If 10.6 is Intel-only, 14 months seems like a pretty reasonable time-frame.
At the very least, it's nice to see he didn't write Q3, which suggests that it won't be like Leopard where we were waiting for ages. Vista came in January and a lot of people were hoping for a release. Then WWDC came and no Leopard. It didn't come out until October.
My concern about a Q1 release is when they talk about it. MacWorld seems a bit early as 14+ months points to March. But then again, they spoke about Leopard in June and released in October.
My concern about a Q1 release is when they talk about it. MacWorld seems a bit early as 14+ months points to March. But then again, they spoke about Leopard in June and released in October.
That time Apple was stilll under the iPhone effect. If I remember correctly, there has been an official announcement explaining the delay.
I don't see however why now Apple would try to release Snow Leopard so early, when we already have official word from the CEO.
I don't see however why now Apple would try to release Snow Leopard so early, when we already have official word from the CEO.
If it's ready, then it's not really early. Jobs only really said it would ship in about a year. This is a pretty vague time-frame. When it comes to software development, it's hard to make an accurate schedule. Usually things are late but sometimes if you do everything right then it's quite acceptable to be ahead of schedule.
I'm not saying its true but SL is supposedly not introducing new features but improving speed and stability.
Why can't it not be ready early next year?
Why is this hard to believe?
I'm not saying its true but SL is supposedly not introducing new features but improving speed and stability.
Why can't it not be ready early next year?
Seeding has not accelerated one bit.
In all of OS X's history, developer seeds begin to accelerate near the end of the development cycle. This has been true of 10.x releases as well as 10.x.x releases. Snow Leopard is not going to be the exception to the rule. This acceleration hasn't even started to happen yet. It usually takes 3-5 months when the seeds start pouring in before OS X 10.x is released. And then it takes about 1 month for the OS to be pressed.
Snow Leopard won't be released before early May...but could take longer if Apple stops its habit of shipping a bug-laden OS (remember 10.4 and 10.5?)
I fully expect Snow Leopard to be released on or after May 2009 and the new frameworks will probably be half-broken until 10.6.3 or 10.6.4.
I'm not being pessimistic, I'm being realistic. This is how it's *always* been and there are no signs that this will change with Snow Leopard.
OK.
Time will tell. Frankly I still don't know why it'll take so long, unless there are *sekrit* features we don't know about that are holding things up.
^^^
OK.
Time will tell. Frankly I still don't know why it'll take so long, unless there are *sekrit* features we don't know about that are holding things up.
Not knowing how extensive the changes in 10.6 are, I can't tell if there are sekrit features that could hold things up. However, I would think OpenCL, Grand Central, and QuickTime X would take a lot of time to test. Unless I was led to believe that these frameworks were much more complex than they are.
Who knows...when I hear things like "rewrite of the Finder" or "rewrite of QuickTime X", I'm sometimes leery. Perhaps some of these aren't really rewrites at all and just Cocoa wrappers and small changes to the code marketed as rewrites.
Still...I haven't seen any significant seeding going on. And until that happens, I'm fairly confident we're still 5+ months away.
I'd love to see 10.6 by March 2009 as much as anyone else...but it's simply not going to happen.
We can all pontificate based on an infinite amount of reasons why SL will be early or late. But the facts are that a maintenance release on this level is unprecedented.
We really don't know how much work went into Grand Central Dispatch, OpenCL and other enhancements prior to Leopards launch. Quicktime X is coming from the iPhone and I've heard that OpenCL contains a large amount of Nvidia CUDA tech.
I just think they should ship it when it's ready. OpenCL is going to take years to deliver power as developers begin to leverage the fast current and future GPU. I'm not looking for overnight changes here.
There really is no reason to rush Snow Leopard. 8-Cores is the max that we have now and there will be few Nehalem based Macs with SMT by the time SL hits.
The interesting thing is that Apple is stating that this is pretty much the end of the road for them putting any significant resources into GCC 4. The future is LLVM and CLANG. I look for a pretty strong Xcode 4 IDE for 10.7.
IThis acceleration hasn't even started to happen yet. It usually takes 3-5 months when the seeds start pouring in before OS X 10.x is released. And then it takes about 1 month for the OS to be pressed.
We still have over 4 months before the end of Q1. Assuming an announcement on January 6th and the builds seeded are ok, it looks like there's enough time. Multiple seeds are only needed when builds are bad and bugs need fixed.
If you consider that Snow Leopard is what Leopard should have been then it's not so much a new development as an extension to their Leopard development schedule. Why would they call it Snow Leopard if it had so many new features? It would be called 10.6 Lion or something.
However, I would think OpenCL, Grand Central, and QuickTime X would take a lot of time to test. Unless I was led to believe that these frameworks were much more complex than they are.
The low-level developments are probably easier to test than user app developments. They will allow for certain programming structures and the code generated by the compilers will always do the same thing. In user apps, developers have to allow for users clicking things in all sorts of orders with multiple preferences and media types etc.
A lot of the tech Apple uses builds on tried and tested software too.
The move to 64-bit will probably be the biggest problem though and should require testing. Unless they can get 32-bit drivers to work seamlessly.
OpenCL is going to take years to deliver power as developers begin to leverage the fast current and future GPU. I'm not looking for overnight changes here.
I think there could be noticeable improvements in the OS Core libraries immediately. One thing that would be nice is fully programmable Core Image shaders. Mixed with Aperture, Final Cut, Motion, it would make for some very powerful real-time effects.
3rd party apps will certainly take longer but there are a lot of CUDA apps whose code can be ported over.
There really is no reason to rush Snow Leopard. 8-Cores is the max that we have now and there will be few Nehalem based Macs with SMT by the time SL hits.
The Mac Pro and Xserves will benefit from Grand Central but Grand Central may also manage threads on the GPU, which have a minimum of 16-cores in the new Macs.
- Apple has to provide complete, extendable, reliable and tested implementations of the new technologies.
- Apple will optimize some patrs of the OS to demonstrate the power of the new techologies.
- Apple will start optimizing their own Apps using new techologies.
I think it does not make a lot of sense if Apple makes only 1. 2 requires that OpenCL and Grand Central Dispatch are very well tested and reliable. They will not start 3 before SL is released, but they may choose to get one application, like iMovie, to showcase the new performance level. But they may get around with QuickTime X only.Also, they have new 64-bit kernel and a lot of drivers to recompile and debug. Some of them will need more than recompile I guess.
I strongly disagree that low-level changes are easier to test than Application level changes. Low level changes affect all applications, sometimes in an unpredictable way. Also, Apple will have to make sure that at least the most popular third party applications are not broken. Otherwise they will get frustrated users, no matter who is responsible for the problem. The best way to achieve this is to leave more time for developers to test their apps after all low level changes are complete.
This makes January release extremely unlikely. My bet is May or later.