You didn't really answer his question.. Personally, i haven't seen any mention of Resolution Independence in either of the 3 seeds of SL so something's fishy here.
Also, in regards to your reply, i'm not sure how it is relevant.
I currently stare at a 22" Dell LCD next to my 15"MBP and if i drag a window and put it in the middle of the 2 screens so that i see half on one side and half on the other, it looks like like the MBP side is about 20% smaller.
RI is supposed to know the size of each pixel on each screen and adjust the size of the window automatically so if i drag it in the middle like i just did, the window will look perfectly equal on either screen. This can ALL be implemented in software as already Apple knows what size the pixel is on both the MBP and the external monitor. I don't see what's taking them soo long to implement this much MUCH needed feature
Adi
Turn it on in Leopard and you'll get an idea of how it requires a bit more effort than knowing the display resolution.
You didn't really answer his question.. Personally, i haven't seen any mention of Resolution Independence in either of the 3 seeds of SL so something's fishy here.
Also, in regards to your reply, i'm not sure how it is relevant.
I currently stare at a 22" Dell LCD next to my 15"MBP and if i drag a window and put it in the middle of the 2 screens so that i see half on one side and half on the other, it looks like like the MBP side is about 20% smaller.
RI is supposed to know the size of each pixel on each screen and adjust the size of the window automatically so if i drag it in the middle like i just did, the window will look perfectly equal on either screen. This can ALL be implemented in software as already Apple knows what size the pixel is on both the MBP and the external monitor. I don't see what's taking them soo long to implement this much MUCH needed feature
Adi
That's not correct. Apple has no idea of the size of your external monitor.
Once upon a time, with Apple's own monitors, it did. The monitors and Apple's cables had pins that told the computer which of Apple's monitors was hooked up, so it knew both the rez and the size.
Today, just as in the past, Apple has no idea what size the screen is for third party devices, only the rez, so it can't equalize the image sizes between them, even if that is what RI is supposed to do, though I've not heard of that.
It's possible that the Mac can tell what size the screen of Apple's monitors are. But all it can do is to make some assumptions as to screen size depending on what the rez is, but it could easily be wrong.
And by the way, rez is the wrong word for total number of pixels.
That's not correct. Apple has no idea of the size of your external monitor.
Once upon a time, with Apple's own monitors, it did. The monitors and Apple's cables had pins that told the computer which of Apple's monitors was hooked up, so it knew both the rez and the size.
Today, just as in the past, Apple has no idea what size the screen is for third party devices, only the rez, so it can't equalize the image sizes between them, even if that is what RI is supposed to do, though I've not heard of that.
It's possible that the Mac can tell what size the screen of Apple's monitors are. But all it can do is to make some assumptions as to screen size depending on what the rez is, but it could easily be wrong.
And by the way, rez is the wrong word for total number of pixels.
I don't agree with that. Apple detected my monitor based on the DDC/E-DDC info. Based on that info it already knows what resolutions it supports and what are the optimal resolutions aka the REAL resolution and chooses it by default. It also know what brand and model it is : DELL 2208WFP
They can easily keep a database with all known monitors where the real pixel size is specified. That db could be updated lets say : once a month. Again, there is no current limitation in implementing Resolution Independence. It's just a matter of getting their APIs up and running. I get the feeling that Cocoa is ready but that anything made with Carbon will fail miserably..
I don't agree with that. Apple detected my monitor based on the DDC/E-DDC info. Based on that info it already knows what resolutions it supports and what are the optimal resolutions aka the REAL resolution and chooses it by default. It also know what brand and model it is : DELL 2208WFP
They can easily keep a database with all known monitors where the real pixel size is specified. That db could be updated lets say : once a month. Again, there is no current limitation in implementing Resolution Independence. It's just a matter of getting their APIs up and running. I get the feeling that Cocoa is ready but that anything made with Carbon will fail miserably..
Adi
It hasn't detected my two Samsungs, or my LaCie. I had to enter them manually.
... I'm posting here since I figured better Snow Leopard knowledge.
IF Apple released a small ARM based dual core netbook with NVIDEO GPU... and they were putting a cut down version of Snow Leopard on it (ie: a new interface, not running regular OSX apps or finder) - is the snow leopard beta currently close to complete for that functionality?
I mean - the kernel would have to be ready along with drivers just for their Netbook. Mail/Calendar would need to be ready for Exchange integration. And some sort of file manager, MobileMe sync, and quicktime would have to be ready (perhaps borrowing quicktime straight off the iPhone?).
Just with the rumour of the 1st quarter release, and so much left to do, I wondered if a Netbook plan for the first release could make sense of it all.
... I'm posting here since I figured better Snow Leopard knowledge.
IF Apple released a small ARM based dual core netbook with NVIDEO GPU... and they were putting a cut down version of Snow Leopard on it (ie: a new interface, not running regular OSX apps or finder) - is the snow leopard beta currently close to complete for that functionality?
I mean - the kernel would have to be ready along with drivers just for their Netbook. Mail/Calendar would need to be ready for Exchange integration. And some sort of file manager, MobileMe sync, and quicktime would have to be ready (perhaps borrowing quicktime straight off the iPhone?).
Just with the rumour of the 1st quarter release, and so much left to do, I wondered if a Netbook plan for the first release could make sense of it all.
So far Apple takes an opposite direction with Snow Leopard: utilizing multi-core and GPU. The only hint for lower level hardware is the advertised installation size reduction. This does mean Apple will not release some new product category but this has nothing to do with the "regular" SL release. Apple released the iPhone prior to SL, right?
... I'm posting here since I figured better Snow Leopard knowledge.
IF Apple released a small ARM based dual core netbook with NVIDEO GPU... and they were putting a cut down version of Snow Leopard on it (ie: a new interface, not running regular OSX apps or finder) - is the snow leopard beta currently close to complete for that functionality?
I mean - the kernel would have to be ready along with drivers just for their Netbook. Mail/Calendar would need to be ready for Exchange integration. And some sort of file manager, MobileMe sync, and quicktime would have to be ready (perhaps borrowing quicktime straight off the iPhone?).
Just with the rumour of the 1st quarter release, and so much left to do, I wondered if a Netbook plan for the first release could make sense of it all.
Apple has one slimmed down version of OS X and that is the iPhone OS for the iPhone and iPod Touch.
What is it with this notion that Apple should splinter engineering efforts and target multiple niche appliance markets?
Apple has to first get the complete system lean, mean and fully Cocoa ready before it has this fantasy of diluting it's efforts to appease the spur of the moment fantasies?
OS's are the most difficult undertaking of all. Being the underpinning for all computer uses, it's got to do the right thing, and do it well.
I remember when NT first came out. It was severely criticized for (among other things) taking up 300 MB of HDD space, and requiring too much RAM, though I've forgotten that number.
Today, we can laugh at those numbers. The OS has become so complex, it's difficult to know how anyone can understand exactly what's in it.
FreeBSD, for example, has code that no one today knows. Code that has been there since the early '60's. Some things may never get straightened out.
Take this with a grain of salt: Spare me the talk about OS's being difficult to undertake. I worked there at NeXT and Apple. You speak as if you're attempting to convey a press release to Wall Street. I am very well aware of the source code, the design goals, the business commitments and much more that was all around me in both companies.
Could you expound a bit more about this? What is it about 10.6 that signifies a "starting point?"? Is it the promotion of Cocoa as the only framework going forward?
I'm looking for more innovation in software coming from Apple. Lately they've hyped their innovation but largely it's been cosmetic IMO. I think they've done some good things in all areas but I'm looking for more positive hits to Clarke's 3rd law.
Any sufficiently advanced technology is indistinguishable from magic
I want to see more WOW stuff.
OS X 10.6 is Cocoa, high end file system support, heavy SMP/OpenMPI support, yada, yada, yada.
The original focus for SMP support, out-of-the-box for Rhapsody was > 32 Processors. We were going to bring an industry leading UNIX system with an even greater ease-of-use to the masses that Apple could never produce, without the merger.
The Carbon transition was 12 months before End of Life. It turned into 12 years where they duplicated masses efforts and time to bring the OS along that truly was a bastardization of Openstep and Mac OS.
Many of the design paradigms from the kernel to the userspace were sacrificed to make the system move forward. Certain steps backward and forward were taken with the XNU Kernel and the best of list from the BSD Filesystem genres.
When OS X 10.6 is finally cleansed you'll understand what I mean by moving forward from where we left off.
Technologies advanced rapidly in the past 12 years and with those additions, the many lessons learned over 2 decades from NeXT and now Apple will finally see the fruits of it's reward.
Steve the pragmatist realized that without a solid growth pattern and heavy pockets the move to Cocoa wasn't going to work like it could at Microsoft; and even they have had massive growing pains for people to move to C#/.NET and not to mention botched work on their Operating Systems.
The latest builds of OS X 10.6 seeds finally releases a GrandCentral API. This means they have nailed down the API, with few minor changes before mass consumption. What it means now is that Developers will have to devote time to Multicore Programming to get the most out of their apps--much of this is managed via Cocoa and thus they won't be reinventing the wheel. The same with OpenCL will be leveraged via Cocoa.
The point of Cocoa isn't to develop the greatest end-to-end API for all the heavy lifting. It's to produce a heavy lifting API for all but the edge cases. Those areas are left up to the developer to address who has that specific set of needs. Over time, those edge cases become more commonly necessary and fold into Cocoa.
With a very productive Developer Community that leverages Services well, applications can delegate portions of their needs to other applications. There isn't a bunch of applications designed to do some of this, some of that and never the two shall mingle.
There is a lot of duplicated functionality in the developer community with very little "services sharing." Cocoa is designed from the ground-up with co-mingling in mind.
I can go on and on but until you see Apple's OS and applications moved to Cocoa, compared to Carbon legacy apps you won't fully understand how it will work.
I can go on and on but until you see Apple's OS and applications moved to Cocoa, compared to Carbon legacy apps you won't fully understand how it will work.
Most excellent response. I kind of figured your response would center around some of the topics mentioned. I for one and glad that Apple has finally made the incumbent mega-developers (Adobe and Microsoft) finally begin to transition their software over to Cocoa.
There is so much potential left untapped in OS X yet it's a fantastic OS today. I expect we'll see a lot of polishing in Snow Leopard and renewed attention to the little features that fallen through the cracks.
I'm also heartened to see Apple finally moving beyond GCC and embracing LLVM and CLANG. At the most recent LLVM Developer Meeting Apple had 6 engineer presenting. I figure by 10.7 they'll be using LLVM 3.0 and CLANG 2.0 and both will be prominent in Xcode 4.5 or Xcode 5.
Take this with a grain of salt: Spare me the talk about OS's being difficult to undertake. I worked there at NeXT and Apple. You speak as if you're attempting to convey a press release to Wall Street. I am very well aware of the source code, the design goals, the business commitments and much more that was all around me in both companies.
Sorry, I wasn't trying to preach to you, though I posted to what you said, it was meant to be more general in nature.
So far Apple takes an opposite direction with Snow Leopard: utilizing multi-core and GPU. The only hint for lower level hardware is the advertised installation size reduction. This does mean Apple will not release some new product category but this has nothing to do with the "regular" SL release. Apple released the iPhone prior to SL, right?
Looking at the new NVIDEA ARM chips (for example) they have 1-4 ARM cores as well as a GeForce GPU on a chip smaller than a penny. I thought Snow Leopard was very focussed on small footprint and efficiency... PLUS the Grand Central etc, though yes I haven't read anything on Snow Leopard for iPhone, AppleTV, or Netbook style devices.
Quote:
Originally Posted by mdriftmeyer
Apple has one slimmed down version of OS X and that is the iPhone OS for the iPhone and iPod Touch.
What is it with this notion that Apple should splinter engineering efforts and target multiple niche appliance markets?
Apple has to first get the complete system lean, mean and fully Cocoa ready before it has this fantasy of diluting it's efforts to appease the spur of the moment fantasies?
If Apple is building OSX for 2 platforms - the iPhone/Touch and the Mac - is there a reason they would build the Mac version and then cut it back, rather than the iPhone/Touch version and then expand?
Whatever Apple's plans, I don't think they'll be jumping on spur of the moment fantasies. If they're working on it then it's a much longer term strategy.
I agree there's nothing to indicate Apple doing this. I just wondered, with reports of Snow Leopard betas having a long way to go but also the Q1 slides, whether Snow Leopard for a specific Apple device with lesser functionality was plausible.
Looking at the new NVIDEA ARM chips (for example) they have 1-4 ARM cores as well as a GeForce GPU on a chip smaller than a penny. I thought Snow Leopard was very focussed on small footprint and efficiency... PLUS the Grand Central etc, though yes I haven't read anything on Snow Leopard for iPhone, AppleTV, or Netbook style devices.
If Apple is building OSX for 2 platforms - the iPhone/Touch and the Mac - is there a reason they would build the Mac version and then cut it back, rather than the iPhone/Touch version and then expand?
Whatever Apple's plans, I don't think they'll be jumping on spur of the moment fantasies. If they're working on it then it's a much longer term strategy.
I agree there's nothing to indicate Apple doing this. I just wondered, with reports of Snow Leopard betas having a long way to go but also the Q1 slides, whether Snow Leopard for a specific Apple device with lesser functionality was plausible.
You do raise an interesting point.
I don't think they could build up from an iPhone/Touch based OS. They could more easily cut down, which is what they've done.
But, with 10.6, they could be looking at future versions in a way that would more easily be moved to both platforms.
I don't know just how modular the OS is now, and whether that's changing in 10.6 in any way that would easy this along.
I don't think they could build up from an iPhone/Touch based OS. They could more easily cut down, which is what they've done.
But, with 10.6, they could be looking at future versions in a way that would more easily be moved to both platforms.
I don't know just how modular the OS is now, and whether that's changing in 10.6 in any way that would easy this along.
mdriftmeyer? Any ideas?
Questions I have that can't be answered:
? Can we expect SL in summer/fall to be on the next Phone, or will that change at some other interval that will be an option for all iPhones or at least iPhone 3Gs? WIll it be free or paid because it's a whole new OS platform?
? Will AppleTV get off of Tiger and bump up to SL to better utilize it's processors? Will it get an Nvidia iGPU?
? Can we expect SL in summer/fall to be on the next Phone, or will that change at some other interval that will be an option for all iPhones or at least iPhone 3Gs? WIll it be free or paid because it's a whole new OS platform?
? Will AppleTV get off of Tiger and bump up to SL to better utilize it's processors? Will it get an Nvidia iGPU?
Nobody try to answer these unanswerable questions!
Will AppleTV get off of Tiger and bump up to SL to better utilize it's processors? Will it get an Nvidia iGPU?
[/INDENT]
The Nvidia ARM cpu has a built in GeForce GPU, and is capable of 720p MP4 playback. Whether Apple uses that, their own design ARM, or a hybrid - I have to think they're considering making an ARM based AppleTV using the same underlying OS as the iPhone.
The Nvidia ARM cpu has a built in GeForce GPU, and is capable of 720p MP4 playback. Whether Apple uses that, their own design ARM, or a hybrid - I have to think they're considering making an ARM based AppleTV using the same underlying OS as the iPhone.
BINGO!
That was exactly my thought Greg save for the Nvidia part. If indeed Apple is the company that just engaged in the multi-year license with Imagination for their PowerVR SGX chip then I can easily see the next ATV being an ARM and PowerVR based unit. Which I also think would be used in some Netbook class device as well in a slightly different configuration.
I certainly think Snow Leopard is going to form the underlying core of even the AppleTV which needs Quicktime X for modern codec playback and to a lesser extent Grand Central Dispatch/OpenCL.
Hell even the future versions of Webkit will loom largely on the ATV because if Apple wants to prevent Flash and Silverlight from taking over the the web they need to show what fast Javascript performance along with HTML5 can do. The ATV should show MobileMe content with superlative performance.
OpenCL and Grand Central are fairly clearly explained by Apple.
QuickTime X, however, worries me a lot. There's supposedly no sign of it in the dev seeds and nobody knows what Apple meant with this gibberish:
Quote:
Using media technology pioneered in OS X iPhone, Snow Leopard introduces QuickTime X, a streamlined, next-generation platform that advances modern media and Internet standards. QuickTime X features optimized support for modern codecs and more efficient media playback, making it ideal for any application that needs to play media content.
Does it mean we'll see H.264/AVC acceleration seen in recent MacBooks? Does it mean QuickTime X will use OpenCL? Does the "streamline" bit mean QuickTime X will finally ditch System 7-era code and QuickTime 1 codecs for "backward compatibility with postage-stamp 5-FPS movies that nobody watches anyway"? Or does "streamline" mean With QuickTime finally take vastly less than 80 megs? Or does it mean it will shed it's godawful "PRO" features and Apple's QT team will finally stop nickel and diming its customers?
What is this "media technology" that was "pioneered in OS X iPhone"...sounds AMAZING! But is it?
For all we know, QuickTime X could simply be H.264/AVC acceleration and with some legacy code stripped out of QuickTime...not something worthy of large version change.
I would hope a large portion of the QT foundation will be rewritten. The rewrite alone with today's technologies and today's frameworks would make QT more "streamlined" and "optimized". But is it being rewritten? Or is it being patched up and marketed as a bleeding-edge media platform?
I hope Apple doesn't blow it...switching the version number from QT7 to QTX better mean that we're going to see something that will blow our socks' off.
Comments
You didn't really answer his question.. Personally, i haven't seen any mention of Resolution Independence in either of the 3 seeds of SL so something's fishy here.
Also, in regards to your reply, i'm not sure how it is relevant.
I currently stare at a 22" Dell LCD next to my 15"MBP and if i drag a window and put it in the middle of the 2 screens so that i see half on one side and half on the other, it looks like like the MBP side is about 20% smaller.
RI is supposed to know the size of each pixel on each screen and adjust the size of the window automatically so if i drag it in the middle like i just did, the window will look perfectly equal on either screen. This can ALL be implemented in software as already Apple knows what size the pixel is on both the MBP and the external monitor. I don't see what's taking them soo long to implement this much MUCH needed feature
Adi
Turn it on in Leopard and you'll get an idea of how it requires a bit more effort than knowing the display resolution.
You didn't really answer his question.. Personally, i haven't seen any mention of Resolution Independence in either of the 3 seeds of SL so something's fishy here.
Also, in regards to your reply, i'm not sure how it is relevant.
I currently stare at a 22" Dell LCD next to my 15"MBP and if i drag a window and put it in the middle of the 2 screens so that i see half on one side and half on the other, it looks like like the MBP side is about 20% smaller.
RI is supposed to know the size of each pixel on each screen and adjust the size of the window automatically so if i drag it in the middle like i just did, the window will look perfectly equal on either screen. This can ALL be implemented in software as already Apple knows what size the pixel is on both the MBP and the external monitor. I don't see what's taking them soo long to implement this much MUCH needed feature
Adi
That's not correct. Apple has no idea of the size of your external monitor.
Once upon a time, with Apple's own monitors, it did. The monitors and Apple's cables had pins that told the computer which of Apple's monitors was hooked up, so it knew both the rez and the size.
Today, just as in the past, Apple has no idea what size the screen is for third party devices, only the rez, so it can't equalize the image sizes between them, even if that is what RI is supposed to do, though I've not heard of that.
It's possible that the Mac can tell what size the screen of Apple's monitors are. But all it can do is to make some assumptions as to screen size depending on what the rez is, but it could easily be wrong.
And by the way, rez is the wrong word for total number of pixels.
That's not correct. Apple has no idea of the size of your external monitor.
Once upon a time, with Apple's own monitors, it did. The monitors and Apple's cables had pins that told the computer which of Apple's monitors was hooked up, so it knew both the rez and the size.
Today, just as in the past, Apple has no idea what size the screen is for third party devices, only the rez, so it can't equalize the image sizes between them, even if that is what RI is supposed to do, though I've not heard of that.
It's possible that the Mac can tell what size the screen of Apple's monitors are. But all it can do is to make some assumptions as to screen size depending on what the rez is, but it could easily be wrong.
And by the way, rez is the wrong word for total number of pixels.
I don't agree with that. Apple detected my monitor based on the DDC/E-DDC info. Based on that info it already knows what resolutions it supports and what are the optimal resolutions aka the REAL resolution and chooses it by default. It also know what brand and model it is : DELL 2208WFP
They can easily keep a database with all known monitors where the real pixel size is specified. That db could be updated lets say : once a month. Again, there is no current limitation in implementing Resolution Independence. It's just a matter of getting their APIs up and running. I get the feeling that Cocoa is ready but that anything made with Carbon will fail miserably..
Adi
I don't agree with that. Apple detected my monitor based on the DDC/E-DDC info. Based on that info it already knows what resolutions it supports and what are the optimal resolutions aka the REAL resolution and chooses it by default. It also know what brand and model it is : DELL 2208WFP
They can easily keep a database with all known monitors where the real pixel size is specified. That db could be updated lets say : once a month. Again, there is no current limitation in implementing Resolution Independence. It's just a matter of getting their APIs up and running. I get the feeling that Cocoa is ready but that anything made with Carbon will fail miserably..
Adi
It hasn't detected my two Samsungs, or my LaCie. I had to enter them manually.
As I said, there's no guarantee.
... I'm posting here since I figured better Snow Leopard knowledge.
IF Apple released a small ARM based dual core netbook with NVIDEO GPU... and they were putting a cut down version of Snow Leopard on it (ie: a new interface, not running regular OSX apps or finder) - is the snow leopard beta currently close to complete for that functionality?
I mean - the kernel would have to be ready along with drivers just for their Netbook. Mail/Calendar would need to be ready for Exchange integration. And some sort of file manager, MobileMe sync, and quicktime would have to be ready (perhaps borrowing quicktime straight off the iPhone?).
Just with the rumour of the 1st quarter release, and so much left to do, I wondered if a Netbook plan for the first release could make sense of it all.
Hey just humour me with this question please...
... I'm posting here since I figured better Snow Leopard knowledge.
IF Apple released a small ARM based dual core netbook with NVIDEO GPU... and they were putting a cut down version of Snow Leopard on it (ie: a new interface, not running regular OSX apps or finder) - is the snow leopard beta currently close to complete for that functionality?
I mean - the kernel would have to be ready along with drivers just for their Netbook. Mail/Calendar would need to be ready for Exchange integration. And some sort of file manager, MobileMe sync, and quicktime would have to be ready (perhaps borrowing quicktime straight off the iPhone?).
Just with the rumour of the 1st quarter release, and so much left to do, I wondered if a Netbook plan for the first release could make sense of it all.
So far Apple takes an opposite direction with Snow Leopard: utilizing multi-core and GPU. The only hint for lower level hardware is the advertised installation size reduction. This does mean Apple will not release some new product category but this has nothing to do with the "regular" SL release. Apple released the iPhone prior to SL, right?
Hey just humour me with this question please...
... I'm posting here since I figured better Snow Leopard knowledge.
IF Apple released a small ARM based dual core netbook with NVIDEO GPU... and they were putting a cut down version of Snow Leopard on it (ie: a new interface, not running regular OSX apps or finder) - is the snow leopard beta currently close to complete for that functionality?
I mean - the kernel would have to be ready along with drivers just for their Netbook. Mail/Calendar would need to be ready for Exchange integration. And some sort of file manager, MobileMe sync, and quicktime would have to be ready (perhaps borrowing quicktime straight off the iPhone?).
Just with the rumour of the 1st quarter release, and so much left to do, I wondered if a Netbook plan for the first release could make sense of it all.
Apple has one slimmed down version of OS X and that is the iPhone OS for the iPhone and iPod Touch.
What is it with this notion that Apple should splinter engineering efforts and target multiple niche appliance markets?
Apple has to first get the complete system lean, mean and fully Cocoa ready before it has this fantasy of diluting it's efforts to appease the spur of the moment fantasies?
OS's are the most difficult undertaking of all. Being the underpinning for all computer uses, it's got to do the right thing, and do it well.
I remember when NT first came out. It was severely criticized for (among other things) taking up 300 MB of HDD space, and requiring too much RAM, though I've forgotten that number.
Today, we can laugh at those numbers. The OS has become so complex, it's difficult to know how anyone can understand exactly what's in it.
FreeBSD, for example, has code that no one today knows. Code that has been there since the early '60's. Some things may never get straightened out.
Take this with a grain of salt: Spare me the talk about OS's being difficult to undertake. I worked there at NeXT and Apple. You speak as if you're attempting to convey a press release to Wall Street. I am very well aware of the source code, the design goals, the business commitments and much more that was all around me in both companies.
Could you expound a bit more about this? What is it about 10.6 that signifies a "starting point?"? Is it the promotion of Cocoa as the only framework going forward?
I'm looking for more innovation in software coming from Apple. Lately they've hyped their innovation but largely it's been cosmetic IMO. I think they've done some good things in all areas but I'm looking for more positive hits to Clarke's 3rd law.
Any sufficiently advanced technology is indistinguishable from magic
I want to see more WOW stuff.
OS X 10.6 is Cocoa, high end file system support, heavy SMP/OpenMPI support, yada, yada, yada.
The original focus for SMP support, out-of-the-box for Rhapsody was > 32 Processors. We were going to bring an industry leading UNIX system with an even greater ease-of-use to the masses that Apple could never produce, without the merger.
The Carbon transition was 12 months before End of Life. It turned into 12 years where they duplicated masses efforts and time to bring the OS along that truly was a bastardization of Openstep and Mac OS.
Many of the design paradigms from the kernel to the userspace were sacrificed to make the system move forward. Certain steps backward and forward were taken with the XNU Kernel and the best of list from the BSD Filesystem genres.
When OS X 10.6 is finally cleansed you'll understand what I mean by moving forward from where we left off.
Technologies advanced rapidly in the past 12 years and with those additions, the many lessons learned over 2 decades from NeXT and now Apple will finally see the fruits of it's reward.
Steve the pragmatist realized that without a solid growth pattern and heavy pockets the move to Cocoa wasn't going to work like it could at Microsoft; and even they have had massive growing pains for people to move to C#/.NET and not to mention botched work on their Operating Systems.
The latest builds of OS X 10.6 seeds finally releases a GrandCentral API. This means they have nailed down the API, with few minor changes before mass consumption. What it means now is that Developers will have to devote time to Multicore Programming to get the most out of their apps--much of this is managed via Cocoa and thus they won't be reinventing the wheel. The same with OpenCL will be leveraged via Cocoa.
The point of Cocoa isn't to develop the greatest end-to-end API for all the heavy lifting. It's to produce a heavy lifting API for all but the edge cases. Those areas are left up to the developer to address who has that specific set of needs. Over time, those edge cases become more commonly necessary and fold into Cocoa.
With a very productive Developer Community that leverages Services well, applications can delegate portions of their needs to other applications. There isn't a bunch of applications designed to do some of this, some of that and never the two shall mingle.
There is a lot of duplicated functionality in the developer community with very little "services sharing." Cocoa is designed from the ground-up with co-mingling in mind.
I can go on and on but until you see Apple's OS and applications moved to Cocoa, compared to Carbon legacy apps you won't fully understand how it will work.
I can go on and on but until you see Apple's OS and applications moved to Cocoa, compared to Carbon legacy apps you won't fully understand how it will work.
Most excellent response. I kind of figured your response would center around some of the topics mentioned. I for one and glad that Apple has finally made the incumbent mega-developers (Adobe and Microsoft) finally begin to transition their software over to Cocoa.
There is so much potential left untapped in OS X yet it's a fantastic OS today. I expect we'll see a lot of polishing in Snow Leopard and renewed attention to the little features that fallen through the cracks.
I'm also heartened to see Apple finally moving beyond GCC and embracing LLVM and CLANG. At the most recent LLVM Developer Meeting Apple had 6 engineer presenting. I figure by 10.7 they'll be using LLVM 3.0 and CLANG 2.0 and both will be prominent in Xcode 4.5 or Xcode 5.
Take this with a grain of salt: Spare me the talk about OS's being difficult to undertake. I worked there at NeXT and Apple. You speak as if you're attempting to convey a press release to Wall Street. I am very well aware of the source code, the design goals, the business commitments and much more that was all around me in both companies.
Sorry, I wasn't trying to preach to you, though I posted to what you said, it was meant to be more general in nature.
So far Apple takes an opposite direction with Snow Leopard: utilizing multi-core and GPU. The only hint for lower level hardware is the advertised installation size reduction. This does mean Apple will not release some new product category but this has nothing to do with the "regular" SL release. Apple released the iPhone prior to SL, right?
Looking at the new NVIDEA ARM chips (for example) they have 1-4 ARM cores as well as a GeForce GPU on a chip smaller than a penny. I thought Snow Leopard was very focussed on small footprint and efficiency... PLUS the Grand Central etc, though yes I haven't read anything on Snow Leopard for iPhone, AppleTV, or Netbook style devices.
Apple has one slimmed down version of OS X and that is the iPhone OS for the iPhone and iPod Touch.
What is it with this notion that Apple should splinter engineering efforts and target multiple niche appliance markets?
Apple has to first get the complete system lean, mean and fully Cocoa ready before it has this fantasy of diluting it's efforts to appease the spur of the moment fantasies?
If Apple is building OSX for 2 platforms - the iPhone/Touch and the Mac - is there a reason they would build the Mac version and then cut it back, rather than the iPhone/Touch version and then expand?
Whatever Apple's plans, I don't think they'll be jumping on spur of the moment fantasies. If they're working on it then it's a much longer term strategy.
I agree there's nothing to indicate Apple doing this. I just wondered, with reports of Snow Leopard betas having a long way to go but also the Q1 slides, whether Snow Leopard for a specific Apple device with lesser functionality was plausible.
Looking at the new NVIDEA ARM chips (for example) they have 1-4 ARM cores as well as a GeForce GPU on a chip smaller than a penny. I thought Snow Leopard was very focussed on small footprint and efficiency... PLUS the Grand Central etc, though yes I haven't read anything on Snow Leopard for iPhone, AppleTV, or Netbook style devices.
If Apple is building OSX for 2 platforms - the iPhone/Touch and the Mac - is there a reason they would build the Mac version and then cut it back, rather than the iPhone/Touch version and then expand?
Whatever Apple's plans, I don't think they'll be jumping on spur of the moment fantasies. If they're working on it then it's a much longer term strategy.
I agree there's nothing to indicate Apple doing this. I just wondered, with reports of Snow Leopard betas having a long way to go but also the Q1 slides, whether Snow Leopard for a specific Apple device with lesser functionality was plausible.
You do raise an interesting point.
I don't think they could build up from an iPhone/Touch based OS. They could more easily cut down, which is what they've done.
But, with 10.6, they could be looking at future versions in a way that would more easily be moved to both platforms.
I don't know just how modular the OS is now, and whether that's changing in 10.6 in any way that would easy this along.
mdriftmeyer? Any ideas?
You do raise an interesting point.
I don't think they could build up from an iPhone/Touch based OS. They could more easily cut down, which is what they've done.
But, with 10.6, they could be looking at future versions in a way that would more easily be moved to both platforms.
I don't know just how modular the OS is now, and whether that's changing in 10.6 in any way that would easy this along.
mdriftmeyer? Any ideas?
Questions I have that can't be answered:
Questions I have that can't be answered:
Nobody try to answer these unanswerable questions!
Nobody try to answer these unanswerable questions!
Will AppleTV get off of Tiger and bump up to SL to better utilize it's processors? Will it get an Nvidia iGPU?
[/INDENT]
The Nvidia ARM cpu has a built in GeForce GPU, and is capable of 720p MP4 playback. Whether Apple uses that, their own design ARM, or a hybrid - I have to think they're considering making an ARM based AppleTV using the same underlying OS as the iPhone.
The Nvidia ARM cpu has a built in GeForce GPU, and is capable of 720p MP4 playback. Whether Apple uses that, their own design ARM, or a hybrid - I have to think they're considering making an ARM based AppleTV using the same underlying OS as the iPhone.
BINGO!
That was exactly my thought Greg save for the Nvidia part. If indeed Apple is the company that just engaged in the multi-year license with Imagination for their PowerVR SGX chip then I can easily see the next ATV being an ARM and PowerVR based unit. Which I also think would be used in some Netbook class device as well in a slightly different configuration.
I certainly think Snow Leopard is going to form the underlying core of even the AppleTV which needs Quicktime X for modern codec playback and to a lesser extent Grand Central Dispatch/OpenCL.
Hell even the future versions of Webkit will loom largely on the ATV because if Apple wants to prevent Flash and Silverlight from taking over the the web they need to show what fast Javascript performance along with HTML5 can do. The ATV should show MobileMe content with superlative performance.
QuickTime X, however, worries me a lot. There's supposedly no sign of it in the dev seeds and nobody knows what Apple meant with this gibberish:
Using media technology pioneered in OS X iPhone, Snow Leopard introduces QuickTime X, a streamlined, next-generation platform that advances modern media and Internet standards. QuickTime X features optimized support for modern codecs and more efficient media playback, making it ideal for any application that needs to play media content.
Does it mean we'll see H.264/AVC acceleration seen in recent MacBooks? Does it mean QuickTime X will use OpenCL? Does the "streamline" bit mean QuickTime X will finally ditch System 7-era code and QuickTime 1 codecs for "backward compatibility with postage-stamp 5-FPS movies that nobody watches anyway"? Or does "streamline" mean With QuickTime finally take vastly less than 80 megs? Or does it mean it will shed it's godawful "PRO" features and Apple's QT team will finally stop nickel and diming its customers?
What is this "media technology" that was "pioneered in OS X iPhone"...sounds AMAZING! But is it?
For all we know, QuickTime X could simply be H.264/AVC acceleration and with some legacy code stripped out of QuickTime...not something worthy of large version change.
I would hope a large portion of the QT foundation will be rewritten. The rewrite alone with today's technologies and today's frameworks would make QT more "streamlined" and "optimized". But is it being rewritten? Or is it being patched up and marketed as a bleeding-edge media platform?
I hope Apple doesn't blow it...switching the version number from QT7 to QTX better mean that we're going to see something that will blow our socks' off.