Apple freezes Snow Leopard APIs as software nears final stretch

[AppleInsider]

Apple this past weekend distributed a new beta of Mac OS X 10.6 Snow Leopard that altered the programming methods used to optimize code for multi-core Macs, telling developers they were the last programming-oriented changes planned ahead of the software's release.

More specifically, Apple is said to have informed recipients of Mac OS X 10.6 Snow Leopard build 10A354 that it has simplified the application programming interfaces (APIs) for working with Grand Central, a new architecture that makes it easier for developers to take advantage of Macs with multiple processing cores.

This technology works by breaking complex tasks into smaller blocks, which are then routed -- or dispatched -- efficiently to a Mac's available cores for faster processing. This allows third-party developers to leverage more of a Mac's hardware resources without having to be well-versed in multithreaded programming.

People familiar with the latest Snow Leopard build say it was these Grand Central "dispatch" methods that were tweaked, or simplified, alongside build 10A354. Going forward, Apple reportedly told developers that "no further API changes are planned for Snow Leopard."

This means developers can now press forward with Snow Leopard versions of their applications with confidence that further Apple-instated changes won't force them to make significant alterations to their code between now and the time the software hits the market. It can also be seen as a sign that the the operating system upgrade is one step closer to reaching a final developmental stretch that will focus on stability and optimization.

With the private release of build 10A354, Apple also reportedly informed developers about a couple of other recent changes to the software, namely the addition of Chinese handwriting recognition support for Macs that include a multi-touch trackpad. Similar software was added to iPhone Software 2.0 a year ago, allowing users to draw Chinese symbols on their handset's touchscreen and then select matching symbols suggested by the iPhone Software.

In support of the handwriting recognition software in Snow Leopard, Apple also reportedly tweaked the system's Language & Text Preference Pane to include support for bidirectional text.

Also drawn to developers attention in build 10A354 is a new codec due to debut with Snow Leopard called MPEG-4 High Efficiency AAC (or HE-AAC). Apple said the codec is an extension of the Low Complexity AAC (or AAC-LC) codec that's optimized for low-bitrate steaming of audio and podcasts.

Mac OS X 10.6 Snow Leopard is expected to hit the market sometime this summer with a near finalized version likely to make a public appearance at Apple's annual developers conference in about a month.

Meanwhile, Apple this weekend also equipped developers with Mac OS X 10.5.7 build 9J61, which corrected one more minor issue with the impending Leopard update.

[ View this article at AppleInsider.com ]

[wessan]

Any sign of new localizations? Bidirectional text support might mean that support for Hebrew and Arabic could be added. I really hope for more European languages support. Absence of built-in Czech localization is a show stopper for a lot of people here in the Czech Republic.

[wwwluckyro]

Could HE-AAC mean AAC+ ? From what I know, the answer is yes. Correct me if I'm wrong.

That means I'll be able to finally listen to my favorite radio stations that stream in that format. Thanks Apple for bringing support for this 5 years old format.

For those interested, more info can be found on this format on it's wikipedia page.

[sphere]

[sorry, deleted]

[solipsism]

This support for Chinese handwriting partly appears to confirm what I've thought for awhile. These new multiple-touch APIs are for Apple's trackpad, not a tablet. I still don't see a viable consumer market for such a device despite the idea being cool. I hope we get trackpad that output data, too, in future Mac notebooks.

Originally Posted by wwwluckyro

Could HE-AAC mean AAC+ ?

I think it's High Efficiency-Advanced Audio Codec for low bandwidth streaming situations. I'll varify that in a little bit.

edit: HE-AAC v1 is also known as AAC+ and AACplus according to Wikipedia.

[melgross]

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

[Hattig]

I presume that HA-AAC / AAC+ support is also coming in iPhoneOS 3 then?

[bloggerblog]

Originally Posted by melgross

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

too bad, i guess Apple had another change of plans!!

[adisor19]

Originally Posted by melgross

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

Ya, so far no trace whatsoever of resolution independence.
More and more i get the feeling that resolution independence can NOT be implemented on Carbon based applications.. so untill Apple dumps Carbon, i doubt we will see RI

I hope i'm wrong and that Apple can pull a rabbit out of its hat for a nice WWDC surprise but i highly doubt it..

Adi

[walshbj]

Originally Posted by solipsism

This support for Chinese handwriting partly appears to confirm what I've thought for awhile. These new multiple-touch APIs are for Apple's trackpad, not a tablet. I still don't see a viable consumer market for such a device despite the idea being cool....

I don't bother with most of the trackpad swipes. I find the keyboard more direct for most stuff. You might be right, but I still think there's a market for a tablet, split between consumers and commercial.

The tablet could catch on all over retail. Apple has gotten industry praise for their POS checkout system - free of cash registers.

The tablet could be great mounted in vehicles that currently have laptops mounted. That's an ergonomic mess. I much rather have a detachable keyboard in a police car, ambulance, fire truck, delivery truck, etc.

I'd like a tablet in my living room as a remote, web browser, etc. Simply a larger and more powerful iPod touch. With an OS devoted to its tablet features.

If I were a doctor carrying a pseudo-tablet laptop around I'd go nuts. They're awkward. Tablet with docking station seems better.

A tablet could redefine thousands of kiosks.

Perfect kitchen counter device.

With a travel keyboard it will appeal to business travelers. The simple act of opening a laptop is a hurdle, ridiculous as it sounds to say so. George Jetson used to complain about having to push a button, his finger throbbed. Things change.

I think there's an expectation of a tablet baked into the stock price right now. If it doesn't show up by October we could see a runoff.

[Booga]

Originally Posted by melgross

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

I seem to be the only one who thinks resolution independence is a bad thing. It consumes vast amounts of processing power in exchange for worse detail rendering.

[MacTel]

So this could really be delivered in final at the end of June or early July then. OSX 10.5.7 is probably the last iteration of Leopard.

With the API frozen and this being an OSX with little consumer wow-factors the rest of the OS seems to be spit and polish in getting done for a June or July release.

[ksec]

Originally Posted by Booga

I seem to be the only one who thinks resolution independence is a bad thing. It consumes vast amounts of processing power in exchange for worse detail rendering.

you are not alone. I too prefer Pixel Perfect Resolution Icons and interface rather then Scaler Graphics.

[adisor19]

Originally Posted by Booga

I seem to be the only one who thinks resolution independence is a bad thing. It consumes vast amounts of processing power in exchange for worse detail rendering.

I don't think you understand what Resolution Independence is on OS X. It doesn't blur and remove detail from anything as the bitmaps required for scaling are available in many resolutions. The rest of the information is vector based which looks good no matter what. Read the Ars Technica review of Leopard to see the mechanism that Apple wants to use in order to achieve Resolution Independence, and you will see that it does NOT make the rendering in worse detail.

The problem seems to be that Carbon applications have too many glitches in order to make it them scale properly.. so it looks like Apple wants to wait till 10.7 or 10.8 to finally get rid of Carbon and enable RI.

Adi

[solipsism]

Originally Posted by MacTel

So this could really be delivered in final at the end of June or early July then. OSX 10.5.7 is probably the last iteration of Leopard.

With the API frozen and this being an OSX with little consumer wow-factors the rest of the OS seems to be spit and polish in getting done for a June or July release.

I think we'll definitely have at least one more point update to Leopard. Apple seems to be waiting on something with 10.5.7 but what that is I could even speculate.

I doubt that SL could arrive any earlier than August with a date more likely late September. The APIs are solidified so developers can start optimizing their apps for GC and such with confidence, but there is still plenty of work that needs to be done to SL, and it's not just adding the rumored Marble UI.

[mjtomlin]

Originally Posted by melgross

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

Supported by Leopard... in fact Safari 4 uses RI controls when zooming.

[solipsism]

Originally Posted by mjtomlin

Supported by Leopard... in fact Safari 4 uses RI controls when zooming.

I don't believe Page Zooming is RI in any way. The text size variance is just the font scaling and the images and page format is scaled to match the text scaling. If the image is bitmap, it will still be bitmap, but if it's a distance vector image it will appear as if RI is there, but only for that image not for the whole display which RI refers to.

[melgross]

Originally Posted by Booga

I seem to be the only one who thinks resolution independence is a bad thing. It consumes vast amounts of processing power in exchange for worse detail rendering.

Computers are plenty powerful enough for that today. Most of the time the computer is sitting around not doing anything. Once it does the render, it doesn't have to do anything more with it until you change whatever you're doing.

I don't know why you would think that there would be bad detail rendering. It would be the same as when enlarging icons now. Perfectly fine. Only pixel mode objects would have that problem, and the idea is to stop using such graphics objects.

[melgross]

Originally Posted by mjtomlin

Supported by Leopard... in fact Safari 4 uses RI controls when zooming.

Not really. Only developers can actually use Rez independence in 10.5 in any useful way, and what Safari does is only what many programs can do already on a very limited scale. Rez independence is a system wide function. It requires every program to adhere to the APIs involved, or there will be serious problems. It's believed, for good reason, that Apple only included it in the developer tools for 10.5 to allow developers play with it and get used to it.

[iVlad]

Originally Posted by MacTel

So this could really be delivered in final at the end of June or early July then. OSX 10.5.7 is probably the last iteration of Leopard.

With the API frozen and this being an OSX with little consumer wow-factors the rest of the OS seems to be spit and polish in getting done for a June or July release.

Not really, Leopard can still be updated up to 10.5.10 even when Snow Leopard is out. This was the case with Tiger.

[macslut]

I can't believe they're finally supporting AAC+. I had all but given up on this...and so did many other content providers. I hope it's not too late for mass adoption.

[Ronbo]

Originally Posted by walshbj

If I were a doctor carrying a pseudo-tablet laptop around I'd go nuts. They're awkward. Tablet with docking station seems better.

There are probably doctors who agree with you. But I'm a doctor who does all of his notes (hospital and office) on a laptop. I do get tired of opening and closing the clamshell. However, Jobs is (unfortunately) right about the bandwidth problem. I type MUCH faster than I write. And I have no desire to use any of the template-based programs. Very few patients' complaints really match the templates... except when you use templates, and suddenly they all do. When I see a note from another doctor, and you can see he/she used one of those programs, I almost never find the information useful to me. The history I get from the patient almost never matches what's in that note.

Now,if Apple makes a tablet, I'll buy one and try VERY hard to find a way to make it work for my practice. (And if it doesn't work for the medical stuff, I'll find something else to use it for and will no doubt love it for that). But there aren't many things that beat a keyboard for getting information into a device quickly (dictations don't work for me; I can do it, but it's much easier to say exactly what you intend if you type it).

My 2 cents. Off-topic, I see. Sorry 'bout that.

[Ronbo]

Originally Posted by melgross

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

Which is very interesting. If ever there had been a release specifically tailored to tidying up specifically this kind of problem, I thought that's what Snow Leopard was all about.

Originally Posted by adisor19

More and more i get the feeling that resolution independence can NOT be implemented on Carbon based applications.. so untill Apple dumps Carbon, i doubt we will see RI

Adi

Someone explain this to me, please. I dabble in Cocoa. I don't do any work in Carbon. Why is Carbon an obstacle to resolution independence?

[melgross]

Originally Posted by Ronbo

There are probably doctors who agree with you. But I'm a doctor who does all of his notes (hospital and office) on a laptop. I do get tired of opening and closing the clamshell. However, Jobs is (unfortunately) right about the bandwidth problem. I type MUCH faster than I write. And I have no desire to use any of the template-based programs. Very few patients' complaints really match the templates... except when you use templates, and suddenly they all do. When I see a note from another doctor, and you can see he/she used one of those programs, I almost never find the information useful to me. The history I get from the patient almost never matches what's in that note.

Now,if Apple makes a tablet, I'll buy one and try VERY hard to find a way to make it work for my practice. (And if it doesn't work for the medical stuff, I'll find something else to use it for and will no doubt love it for that). But there aren't many things that beat a keyboard for getting information into a device quickly (dictations don't work for me; I can do it, but it's much easier to say exactly what you intend if you type it).

My 2 cents. Off-topic, I see. Sorry 'bout that.

I agree completely. It's always amusing, when it isn't frustrating, when a form requires you to check off "always, sometimes, never", when it's really almost never, or almost always. We recently had that problem with my daughter. We had her going to The NYU Child Development Center to get a final evaluation of her ADD, which has become milder as she has gotten older. The 6 page form had us scratching our heads at times. Often, the answer is "not quite, but maybe almost one of the above, but maybe not".

You get forced into a position where your responses aren't quite correct.

[melgross]

Originally Posted by Ronbo

Which is very interesting. If ever there had been a release specifically tailored to tidying up specifically this kind of problem, I thought that's what Snow Leopard was all about.

I thought so too. I had my hopes up.

[Trajectory]

Originally Posted by macslut

I can't believe they're finally supporting AAC+. I had all but given up on this...and so did many other content providers. I hope it's not too late for mass adoption.

I just hope this doesn't mean we'll have to shell out more money to upgrade our libraries to the new AAC+ format like Apple did with the iTunes Plus format.

[solipsism]

Originally Posted by Trajectory

I just hope this doesn't mean we'll have to shell out more money to upgrade our libraries to the new AAC+ format like Apple did with the iTunes Plus format.

iTunes Plus was a doubling of bitrate and removal of DRM, it wasn't a different audio codec. This audio codec is designed for low bitrate streaming for low bandwidth situations. You won't find this on the iTS for music.

[Garamond]

Originally Posted by melgross

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

In reality, what are the benefits of resolution independence? Would it mean so much to the general Mac user as people tend to believe?

[MotherBrain]

Originally Posted by melgross

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

Man, I sure hope so, my eyes are getting too old to be squinting at OS X through high-density monitors. Windows does this without any issues. One other thing that I really, really wish OS X had for the longest time, and it's so damn simple, is the ability to resize windows from anywhere on the window instead of just the stupid little corner on the bottom right. Does anyone know if we will finally have this in SL? I thought we were supposed to have that in 10.5?!?

It's these 2 issues that bug me the most about having to work cross-platform.

[MotherBrain]

Originally Posted by Booga

I seem to be the only one who thinks resolution independence is a bad thing. It consumes vast amounts of processing power in exchange for worse detail rendering.

Windows does this just fine, at least Vista anyway, don't know about XP. But there is no performance hit whatsoever.

[solipsism]

Originally Posted by MotherBrain

Windows does this just fine, at least Vista anyway, don't know about XP. But there is no performance hit whatsoever.

Windows doesn't utilize real RI, they use Windows Presentation Foundation API which, to put it simply, is a graphical subsystem for rich user interface development. While not true RI I wish Apple would offer a better option for scaling their GUI if they aren't going to ready RI anytime soon. Windows has always handled this better then OS X.

[macxpress]

I wasn't aware Windows Vista has RI???

[shadow]

Originally Posted by melgross

This looks to be killing the possibility for us seeing rez independence in 10.6. That would require other API's, and if we're not seeing it in the new developer builds, then, I assume it's out. Too bad.

All the APIs needed are available since Tiger. There is nothing special in this. The most important part is to follow a number of rules which boil down to the following: don't assume the pixel size and actual size are the same. Then, the only APIs needed are those to align the image to an integer pixel boundary to avoid blur. The tricky part is to align several UI elements in the same way when needed.

[lilgto64]

Originally Posted by MotherBrain

Man, I sure hope so, my eyes are getting too old to be squinting at OS X through high-density monitors. Windows does this without any issues. One other thing that I really, really wish OS X had for the longest time, and it's so damn simple, is the ability to resize windows from anywhere on the window instead of just the stupid little corner on the bottom right. Does anyone know if we will finally have this in SL? I thought we were supposed to have that in 10.5?!?

It's these 2 issues that bug me the most about having to work cross-platform.

resizing from somewhere other than the bottom right corner would be nice - especially when the bottom right corner is off the bottom of the screen and you cannot get to it - though at the other end of the extreme I find Windows to be aggravating when I accidentally resize windows without meaning to.

There are some third party options that claim to allow window resizing from anywhere or with keyboard shortcuts - I have not tried any of them myself but if you go to versiontracker.com and search on "resize window" you will get some shareware and freeware hits.

You can also try the keyboard combo - Command-Option-8 plus Command-Option = or Command-Option - to turn on Zoom then to zoom in and out - although that does not resize the window - it does zoom the screen contents which could be useful in some cases (I use this on the computer attached to my TV - at 1920x1080 on a 65" screen most of the time it is okay - but sometimes I need to zoom in a bit to get a better look at what I am doing.

[melgross]

Originally Posted by Garamond

In reality, what are the benefits of resolution independence? Would it mean so much to the general Mac user as people tend to believe?

It depends. I've got a 24" monitor. The menu bar on the top is thin, and the type small. But I can still use it. On a much higher rez 30" monitor, it's verging on the edge of illegibility. The same thing is true for icons, type etc. For programs, we get the same problems. With Rez independence, the entire screen can be brought up in size the way you could do it with CRT monitors, but with all the resolution your monitor has. Things will still be sharp, and you might even see a bit more detail as a result of vectors being used instead of bitmaps.

You know where in a program you can magnify the page you're working on, but not the program itself? Well, with Rez independence, everything will be larger. If your eyesight isn't the best, this will be a godsend.

[melgross]

Originally Posted by MotherBrain

Man, I sure hope so, my eyes are getting too old to be squinting at OS X through high-density monitors. Windows does this without any issues. One other thing that I really, really wish OS X had for the longest time, and it's so damn simple, is the ability to resize windows from anywhere on the window instead of just the stupid little corner on the bottom right. Does anyone know if we will finally have this in SL? I thought we were supposed to have that in 10.5?!?

It's these 2 issues that bug me the most about having to work cross-platform.

Windows doesn't do Rez independence. It doesn't do any more than OS X does now. There are various ways to enlarge what's on the screen, but Apple's Rez independence prototype is the best around right now.

Resizing windows has nothing to do with Rez independence either. Sometimes having the ability to drag any side is good, sometimes it doesn't matter.

[melgross]

Originally Posted by MotherBrain

Windows does this just fine, at least Vista anyway, don't know about XP. But there is no performance hit whatsoever.

No, it does not.

[melgross]

Originally Posted by shadow

All the APIs needed are available since Tiger. There is nothing special in this. The most important part is to follow a number of rules which boil down to the following: don't assume the pixel size and actual size are the same. Then, the only APIs needed are those to align the image to an integer pixel boundary to avoid blur. The tricky part is to align several UI elements in the same way when needed.

The special thing is preparing all the elements needed for RI. That's one. The second it to get all the thousands of developers on board. That's not so easy.

[foljs]

Originally Posted by adisor19

Ya, so far no trace whatsoever of resolution independence.
More and more i get the feeling that resolution independence can NOT be implemented on Carbon based applications.. so untill Apple dumps Carbon, i doubt we will see RI Adi

Resolution independence is a long term goal... We won't be seeing it any time soon.

As a matter of fact, you don't need any new APIs for RI. You can already (as in 10.5) tweak some developer preferences to enable it. But the view with zoomed windows won't be pretty (icons and stuff not aligned, low-res bitmaps pixelized, etc)...

Carbon might be a part of this, but the main reason is that it needs changes from all software developers too. Current programs are not written to be RI. Not even all of Apple's programs are.

[foljs]

Originally Posted by Booga

I seem to be the only one who thinks resolution independence is a bad thing. It consumes vast amounts of processing power in exchange for worse detail rendering.

You confuse "resolution independence" with changing resolutions.

Actually RI does not consume vast amounts of processing power (it's all in the GPU, and it's not a complicated process).

And the detail rendering is MUCH, MUCH improved.

It's only when currently changing resolution on an LCD/TFT screen to something other than the native that you get worse rendering. And this worse rendering is because of lack of RI.