[SoClose]

Originally Posted by cambro

100% correct.

This spells the end of Mac OS application development. No company in their right mind would EVER make an OS X application now.

Huh? There are plenty of companies making some pretty good money off Mac users, why would they want to stop?

Originally Posted by cambro

We will all be running Windows apps cobbled onto the Apple OS...which itself will be hacked to run on Dells and Gateways.

Dude, that's a pretty huge leap. You're forgetting about the applications that have drawn many new customers to the Macintosh: iLife '05.

Originally Posted by cambro

The sad part of this is that the ONLY positive outcome of a switch to a platform that is virtually identical to Windows machines is rapid market share growth of the OS.

What?! Apple is replacing the CPU of the Macintosh computer, not rehosting Mac OS X on the Windows kernel! The platforms are still very different. It doesn't matter that the processor is the same, it's all about the APIs. Cripes Cisco uses PowerPC chips in their routers! OMFG -- my Mac is now a network router! ARGHHHHH!

Apple has an amazing developer productivity proposition in the Cocoa APIs and related frameworks. Have you seen what modern C++ Windows development is like? Almost worse is C# development! Apple has a sliver of the developers that Microsoft has and can produce incredibly high-quality software on a regular cycle with small developer teams. That's reason for many ISVs to sit-up and take notice. Now that Apple is on Intel, I expect more enterprise developers to come Apple's way, not less.

Originally Posted by cambro

This, of course would spell the end of what, to me anyway, makes Apple unique (no viruses, integrated mac-like solutions etc.).

Does your brain ever talk to your fingers? The security problems on Windows are just that -- they're because of Windows not the the CPU. You could run Windows XP on PowerPC chips (which you can do with Windows XP Embedded) and it'll still be suspectible to the same security problems as Windows XP on Intel. It's the OS that's insecure, not the chips. In fact, Intel has been aggressively building in hardware-based security features. In fact, the Macintosh now has the potential to become even more secure in 2006!

Originally Posted by cambro

In short, Apple will no longer become a hardware company and will become Microsoft 2.

Bzzzt. Wrong. Dead wrong. They certainly won't be Microsoft 2. Apple will continue to do what Apple does best: out innovate it's competitors. Innovation will lead to more customers. More customers for us to welcome into the Mac community!

[Detrius]

Originally Posted by Cadaver

I'd like to see something like the Citrix/Metaserver solution Windows has for implementing virtual machine access. In case you're not familiar with it, its a lightweight Windows app that lets you use an application hosted on a remote machine as if it was installed on your machine. For example, at work we use Microsoft Outlook for email, scheduling, etc. From my PC at home, I can log in and use Outlook (hosted on the office's servers) just like I would use any other Windows app - its own window, menu bar, etc... Now, its not just logging in and using using a copy of Outlook loaded on my machine - I don't have Outlook installed on my machine. I can do this with any app hosted on my work's servers (Word, PowerPoint, etc.) as if they were installed locally. Think rootless X11 apps on MacOS X.

I'd love for something like this to be available on MacOS X (even MacOS X Server). A true remote GUI.

That is precisely what I was talking about. I was just trying to describe it in fewer words that people would understand who don't know about Citrix or X11's capabilities. (I still can't believe MS beat Apple on this one. Apple has Unix!)

However, I don't particularly care if it's fullscreen/windowed (think VPC or VNC) or rootless... I guess I would prefer both, as each has it's positives and negatives depending on the situation.

[SoClose]

Originally Posted by Detrius

That is precisely what I was talking about. I was just trying to describe it in fewer words that people would understand who don't know about Citrix or X11's capabilities. (I still can't believe MS beat Apple on this one. Apple has Unix!)

However, I don't particularly care if it's fullscreen/windowed (think VPC or VNC) or rootless... I guess I would prefer both, as each has it's positives and negatives depending on the situation.

I do this today and it's a huge benefit. I connect to our Linux servers via ssh and can remotely execute rootless applications (for those apps that don't have a web front-end).

[godzookie2k]

I use citrix on my mac to access our timesheet software at work. Works well enough.

[cambro]

Originally Posted by SoClose

Huh? There are plenty of companies making some pretty good money off Mac users, why would they want to stop?

I agree with most of your points, at least in part. I was painting a purposefully antagonistic portrait of what this change means.

However, what I was mainly trying to question is why any deveoper in their right mind would even bother with Cocoa/Carbon when, in all likelihood, pure Windows apps will be made to run fine with little or no speed hits under OS X. It just makes NO FINANCIAL SENSE. You can bet Adobe lays off a lot of the mac development staff in the coming years, for example.

After I posted this comment, Slashdot developers are talking about this very question (slashdot thread).

So, I'd say it may not be as laughable as you may (want) to think.

The most important point is that the distinction between the Mac and Windows platform has been significantly and fundamentally blurred...not necessarily to the end-user of the KILLER OS we enjoy as mac users...but to developers of mac software.

[SoClose]

Originally Posted by cambro

I agree with most of your points, at least in part. I was painting a purposefully antagonistic portrait of what this change means.

However, what I was mainly trying to question is why any deveoper in their right mind would even bother with Cocoa/Carbon when, in all likelihood, pure Windows apps will be made to run fine with little or no speed hits under OS X. It just makes NO FINANCIAL SENSE. You can bet Adobe lays off a lot of the mac development staff in the coming years, for example.

After I posted this comment, Slashdot developers are talking about this very question (slashdot thread).

So, I'd say it may not be as laughable as you may (want) to think.

The most important point is that the distinction between the Mac and Windows platform has been significantly and fundamentally blurred...not necessarily to the end-user of the KILLER OS we enjoy as mac users...but to developers of mac software.

I didn't mean to be so overly antoganistic. However, here are some quotes from that Slashdot article:

"...developing on the Mac platform is the best its ever been..."

"I think this is the perfect time to start developing Mac software..."

I think many developers are viewing this as a positive rather than a negative. As I said before, Apple has a very strong story for developer productivity. As more developers produced well-factored applications Apple is in a great position to pitch the notion of creating a Cocoa front-end to applications that communicate with non-UI processes or services.

[besson3c]

Originally Posted by Big Mac

But what about dual-booting? To wit:

Customer: "Do you still offer a Mac version of application x?"

Sales: "You can now run our Windows version on your Mac by booting into Windows."

Customer: "But I don't want to run Windows."

Sales: "Well, I understand your concern, but our company decided that it was more economical to support the Windows platform solely. And there's no reason why can't run our Windows software on your Mac now that it is Intel based."

Windows will run on anything and everything. Apple said there will not be an effort made to prevent it. And then the final developer exodus will commence. Companies will drop Mactel support (and that perpetually troublesome Mac support with it) as soon as these "Macs" are capable of running Windows. Please, show me how I'm wrong.

1) It requires Mac users to have purchased a Windows license

2) It requires Mac users to have purchased or downloaded the software necessary that will provide them with the Windows layer

3) It requires the company to support this software, if they are claiming Mac support

4) If the software is too geek centric, this company may alienate their target audience if their target audience is not soley geek

5) It is doubtful that WINE would offer integration such as copy and paste, and certainly things like the menu placement would be inconsistent with the Mac experience. Like X11 and Classic, this is an awkward solution (but better than nothing). If the company is claiming Mac support, they'll have to deal with this.

6) All of this could change the definition of what "Mac support" really is. I suspect that claiming Mac support through this layer is not going to sustain a company without serious legal or PR backlash.

My prediction is that companies who had no intention to support the Mac anyway can offer this as a solution to their customers. It will always be a crippled solution, but it will get us closer than we would have been otherwise.

[Eriamjh]

Native OS9 Booting!

[Nebrie]

Some companies such as Corel have done so. They announced Linux versions of their products. When the users installed the software, they found out it was nothing more than the Windows version with wine.

[Chuckit]

Linux hardly has the UI standards that OS X does.

[SMacTech]

Originally Posted by cambro

100% correct.

This spells the end of Mac OS application development. No company in their right mind would EVER make an OS X application now.

.

I see, so for the next two years while PPC is still in production and the many millions that still have PPC/OSX and then on Intel running Rosetta or native intel/OSX we STOP making apps today, right ?

Ok, I will let my wife know why I stopped writing software that SELLS !!!!

[Simon]

Originally Posted by Detrius

Back to the issue of 10.5...

I hope they finally get their butts in gear and offer a remote graphical terminal. Why shouldn't we be able to share one very powerful Macintosh amongst several people with far less powerful Macs? At least they offer a cheap Mac now to fulfill the terminal end.

And along those lines, it would be really freaking awesome if you could connect two keyboards, two mice, and two monitors to one high-end Mac and have two people sit down at it and use it. They could even offer us a dual multi-core machine with the capability to hook up upwards of 4 to 8 displays and keyboards, and then we would have a true workstation... It would be optimal if we could hook up more displays than keyboards so that some people could have multiple monitors.

drool...

If Apple would do it, I doubt they would advertise it as the key feature of Leopard (like Spotlight or Dashboard under Tiger), but this is the one feature that would get me to spend $129 in a second. I can't think of another bit of functionality I've missed more under OS X. It's a feature my colleagues and I use daily under Linux and to think that I can even do it with that other OS from Redmond, but not with my beloved Mac is really about as bad as it gets.

[mishakim]

Originally Posted by godzookie2k

I use citrix on my mac to access our timesheet software at work. Works well enough.

what we want is for the Mac server to provide similar functionality - so you can run Mac software that only lives on the server, from your client (maybe even a non-mac client). Or even software that you might happen to have on the client, but would run faster on the server, or would have access to different resources.

Though it would also be nice to have rootless Citrix windows, like XP does. That's for Citrix to implement though, not Apple. If there's anything in Tiger holding them back, that would be something nice for Leopard to resolve.

[ZinkDifferent]

Originally Posted by SoClose

Well, I guess we now know the fate of the rumored "Yellow Box" version of the Cocoa APIs that would run on Windows/Intel. It was there the whole time! Can you imagine being the Apple engineers walking around with the secret that Mac OS X has been compiled and executing on the Intel platform for almost 5 years?!

It's not exactly a secret if everyone has known about it for 5 years -- well, everyone but you, it seems.

Seriously, almost everyone at Apple would have confirmed to you, anytime during the past years, that every version of OS X had been running on intel, internally at Apple. Q/Q and development reasons, you know.

ZinkDifferent

[powerdafuture]

freakin OGG vorbis support in iTunes...is it too much to ask for?

[krove]

Originally Posted by mishakim

what we want is for the Mac server to provide similar functionality - so you can run Mac software that only lives on the server, from your client (maybe even a non-mac client). Or even software that you might happen to have on the client, but would run faster on the server, or would have access to different resources.

Though it would also be nice to have rootless Citrix windows, like XP does. That's for Citrix to implement though, not Apple. If there's anything in Tiger holding them back, that would be something nice for Leopard to resolve.

NetBooting provides this functionality, but I think it is an all-or-none solution. All Apple would need to do it add on support for serving Applications off a server similar to netbooting without serving the whole operating system. Right?

[krove]

I think we will see cocoa APIs to tap into the new meta-data, low-level system calls that Apple introduced with Tiger. We'll see implementations from all the iLife apps that save meta-data into the space provided by these system calls--data that is currently not in the file, like data that currently resides in the iPhoto database like title, comment, etc.

Maybe we'll actually see a Cocoa-based finder, or at least an implementation that doesn't suck so much. I hate how the finder windows cannot remember their sizes, positions, column widths, etc. I want a finder window that has a properly-sized side bar with three wide columns showing. You would think that is simple enough, but my Finder cannot remember how I setup my windows!

A modern AFP (apple file protocol). The current one just sucks, is veeerrrry slow, and is not intuitively integrated into the Finder. I think Apple could do better. Rendezvous, err, bonjour-based file protocol, anyone? We need something light, fast, and easy to use (but powerful, too).

Mobile home directories. I want a single home directory for all my computers. Surely that cannot be that difficult to implement? Or maybe it is, what with user permissions and all....but I'm sure Apple can think up something ingenious to work around that *small* issue.

If not mobile home directories, then a synchronization GUI that keeps ALL my preferences etc in my home folder synced between two or more Macs. I do not want some .Mac inbetween either. Direct, Mac to Mac, home directory synchronization.

I imagine we'll start to see the fruits of Apple's resolution-independent GUI underpinnings that showed up in Quartz Debug in Tiger. Not a full-fledged move to resolution-independent display output, but a modest start. Along with this, Apple may provide for some alternate formats for current GUI bitmaps: PDF based icons, etc. Quartz is built on PDF, so this should be a no-brainer.

Well, that's all I could come up with for now. I'm sure others must have some other inspiring ideas, too.

[Chuckit]

Originally Posted by krove

I think we will see cocoa APIs to tap into the new meta-data, low-level system calls that Apple introduced with Tiger.

What metadata functions aren't available through Cocoa right now?

Originally Posted by krove

We'll see implementations from all the iLife apps that save meta-data into the space provided by these system calls--data that is currently not in the file, like data that currently resides in the iPhoto database like title, comment, etc.

Isn't this data already indexed by Spotlight?

[msuper69]

Originally Posted by SoClose

Huh? There are plenty of companies making some pretty good money off Mac users, why would they want to stop?



Dude, that's a pretty huge leap. You're forgetting about the applications that have drawn many new customers to the Macintosh: iLife '05.



What?! Apple is replacing the CPU of the Macintosh computer, not rehosting Mac OS X on the Windows kernel! The platforms are still very different. It doesn't matter that the processor is the same, it's all about the APIs. Cripes Cisco uses PowerPC chips in their routers! OMFG -- my Mac is now a network router! ARGHHHHH!

Apple has an amazing developer productivity proposition in the Cocoa APIs and related frameworks. Have you seen what modern C++ Windows development is like? Almost worse is C# development! Apple has a sliver of the developers that Microsoft has and can produce incredibly high-quality software on a regular cycle with small developer teams. That's reason for many ISVs to sit-up and take notice. Now that Apple is on Intel, I expect more enterprise developers to come Apple's way, not less.



Does your brain ever talk to your fingers? The security problems on Windows are just that -- they're because of Windows not the the CPU. You could run Windows XP on PowerPC chips (which you can do with Windows XP Embedded) and it'll still be suspectible to the same security problems as Windows XP on Intel. It's the OS that's insecure, not the chips. In fact, Intel has been aggressively building in hardware-based security features. In fact, the Macintosh now has the potential to become even more secure in 2006!



Bzzzt. Wrong. Dead wrong. They certainly won't be Microsoft 2. Apple will continue to do what Apple does best: out innovate it's competitors. Innovation will lead to more customers. More customers for us to welcome into the Mac community!

[McKenna]

Originally Posted by SoClose

Well, I guess we now know the fate of the rumored "Yellow Box" version of the Cocoa APIs that would run on Windows/Intel. It was there the whole time! Can you imagine being the Apple engineers walking around with the secret that Mac OS X has been compiled and executing on the Intel platform for almost 5 years?!

From August 2002: http://www.eweek.com/article2/0,1759,1656622,00.asp

--Tom

[bowwowman]

I free'd my mind... now it won't come back

if you loved while it was there, it WILL come back.......
someday hehehe

[WJMoore]

Originally Posted by Chuckit

What metadata functions aren't available through Cocoa right now?

This meta data:

The final, and perhaps most important limitation of Tiger's extended attribute implementation can also be seen as a feature. The extended attribute APIs exist only at the BSD layer. There are no Carbon, Cocoa, or even Core Foundation interfaces to them at this time. If an application wants to use extended attributes, it has to use the (decidedly low-level) BSD APIs.

So whilst you could use extended attributes in a Cocoa app now you have to resort to low(er)-level calls instead of being able to do it with a nice object oriented Cocoa interface.

[Miniryu]

Originally Posted by Big Mac

Please, show me how I'm wrong.

#1. Windows costs ~$400, Mac OS costs ~$150
#2. Longhorn's promise isn't looking too hot at this point. When it is released head to head with Leopard and it can't hang, there won't be as many sales people pushing it. Even if they do, the customer will have the argument that they use the better operating system.

Seriously, the big Apple vs. M$ showdown was supposed to be Longhorn vs. Tiger. M$ knew they would lose this round, so they backed out and now they have to face Leopard.

[vsurfer]

Originally Posted by Simon

If Apple would do it, ... but this is the one feature that would get me to spend $129 in a second. I can't think of another bit of functionality I've missed more under OS X. It's a feature my colleagues and I use daily under Linux and to think that I can even do it with that other OS from Redmond, but not with my beloved Mac is really about as bad as it gets.

I'm with you on that . . . but I probably wouldn't need to until 2007, when the oldest macs in the house become truly useless. For now there's a G5 imac, a G4 powerbook and even a G3 /400 iMac which is still holding its end up admirably (which is pretty amazing in of itself -- it's happy enough with panther inside).

[clarkgoble]

Originally Posted by WJMoore

This meta data:
So whilst you could use extended attributes in a Cocoa app now you have to resort to low(er)-level calls instead of being able to do it with a nice object oriented Cocoa interface.

Although let's be honest. Writing an Obj-C wrapper for Cocoa (or for Carbon, a C++) around these calls would be fairly trivial.

[jamil5454]

Originally Posted by alphasubzero949

I'm sure that his Steveness had a bounty over their heads if they leaked that secret.

Considering they've had the whole operating system (apart from the gui) ported to x86 ever since Darwin became open source, I'm not too surprised by the Intel announcement. OS X seems to be equally designed for both architectures simultaneously.

[moki]

Originally Posted by cambro

However, what I was mainly trying to question is why any deveoper in their right mind would even bother with Cocoa/Carbon when, in all likelihood, pure Windows apps will be made to run fine with little or no speed hits under OS X. It just makes NO FINANCIAL SENSE. You can bet Adobe lays off a lot of the mac development staff in the coming years, for example.

Have you seen the demos of CoreImage, CoreVideo, etc?

Apple is spending an awful lot of time and energy making sure that their APIs allow developers to do things that couldn't be done before, quickly and easily. It is not a stretch to believe that some of the more innovative software in terms of functionality and user experience will be built on top of MacOS X Tiger and beyond.

[ShotgunEd]

Originally Posted by krove

Maybe we'll actually see a Cocoa-based finder, or at least an implementation that doesn't suck so much. I hate how the finder windows cannot remember their sizes, positions, column widths, etc. I want a finder window that has a properly-sized side bar with three wide columns showing. You would think that is simple enough, but my Finder cannot remember how I setup my windows!

The Finder does remember window position and size of columns etc, it just that you have to do some very unintuitive closing and reopening of windows.

A modern AFP (apple file protocol). The current one just sucks, is veeerrrry slow, and is not intuitively integrated into the Finder. I think Apple could do better. Rendezvous, err, bonjour-based file protocol, anyone? We need something light, fast, and easy to use (but powerful, too).

I agree with this, OSX's networking sucks.

Mobile home directories. I want a single home directory for all my computers. Surely that cannot be that difficult to implement? Or maybe it is, what with user permissions and all....but I'm sure Apple can think up something ingenious to work around that *small* issue.

I think this is a feature (or certainly a similar one exists) in OSX Server. So technological limitation isn't the reason we don't have it. But I agree again, It'd be nice to have the option of either synced home directories, or some method of sharing them. With advances in WiFi (I forget the name of the version that intel is developing) it'd be sweet to have a home directory on your desktop mac that was available yo all your other macs wirelessly.

I imagine we'll start to see the fruits of Apple's resolution-independent GUI underpinnings that showed up in Quartz Debug in Tiger. Not a full-fledged move to resolution-independent display output, but a modest start. Along with this, Apple may provide for some alternate formats for current GUI bitmaps: PDF based icons, etc. Quartz is built on PDF, so this should be a no-brainer.

I see this happening pretty soon. If Apple keep bringing out larger and larger displays its gonna get to the point where you can't see the menubar. A scalable UI removes the limitation and screen sizes can get silly.

[LeeG]

Is there any chance, much like 'carbon' apps, that someone could write a compiler that would allow 1 binary to open both in OSX-Intel and Wintel?

It seems if someone could write a compiler that would allow this (as carbon did), that even MORE developers would write for mac - as they could output one program that ran on both.

Now I realize thats a lot harder than it seems, esp with different APIs for each OS, and also that apps would lose their "mac-ness" if they weren't developed specifically for macs, and to take advanatge of mac-only features, but it would be better than being told to reboot with windows and use the windows version....

L

[pat++]

I would like to see :

- NEW Finder that actually works and is not dead slow.
- read/write FTP from Finder. C'mon, Microsoft is doing this since Win95 !
- Get rid of this mess concerning UI look. One (or two) themes... NOT FOUR!
- Unification of Splotlight/Finder search window... why can't I specify search criteriras in Spotlight window???

[wnuez]

Originally Posted by krove

Mobile home directories. I want a single home directory for all my computers. Surely that cannot be that difficult to implement? Or maybe it is, what with user permissions and all....but I'm sure Apple can think up something ingenious to work around that *small* issue.

If not mobile home directories, then a synchronization GUI that keeps ALL my preferences etc in my home folder synced between two or more Macs. I do not want some .Mac inbetween either. Direct, Mac to Mac, home directory synchronization.

they might be on their way. have a look at: System->Library->CoreServices->Menu Extras->HomeSync.menu

[msuper69]

Originally Posted by SoClose

Huh? There are plenty of companies making some pretty good money off Mac users, why would they want to stop?



Dude, that's a pretty huge leap. You're forgetting about the applications that have drawn many new customers to the Macintosh: iLife '05.



What?! Apple is replacing the CPU of the Macintosh computer, not rehosting Mac OS X on the Windows kernel! The platforms are still very different. It doesn't matter that the processor is the same, it's all about the APIs. Cripes Cisco uses PowerPC chips in their routers! OMFG -- my Mac is now a network router! ARGHHHHH!

Apple has an amazing developer productivity proposition in the Cocoa APIs and related frameworks. Have you seen what modern C++ Windows development is like? Almost worse is C# development! Apple has a sliver of the developers that Microsoft has and can produce incredibly high-quality software on a regular cycle with small developer teams. That's reason for many ISVs to sit-up and take notice. Now that Apple is on Intel, I expect more enterprise developers to come Apple's way, not less.



Does your brain ever talk to your fingers? The security problems on Windows are just that -- they're because of Windows not the the CPU. You could run Windows XP on PowerPC chips (which you can do with Windows XP Embedded) and it'll still be suspectible to the same security problems as Windows XP on Intel. It's the OS that's insecure, not the chips. In fact, Intel has been aggressively building in hardware-based security features. In fact, the Macintosh now has the potential to become even more secure in 2006!



Bzzzt. Wrong. Dead wrong. They certainly won't be Microsoft 2. Apple will continue to do what Apple does best: out innovate it's competitors. Innovation will lead to more customers. More customers for us to welcome into the Mac community!

Finally, a reply from someone who gets it. Too many post based on emotion rather than logic.
Great post.

[msuper69]

Originally Posted by Miniryu

#1. Windows costs ~$400, Mac OS costs ~$150
#2. Longhorn's promise isn't looking too hot at this point. When it is released head to head with Leopard and it can't hang, there won't be as many sales people pushing it. Even if they do, the customer will have the argument that they use the better operating system.

Seriously, the big Apple vs. M$ showdown was supposed to be Longhorn vs. Tiger. M$ knew they would lose this round, so they backed out and now they have to face Leopard.

(emphasis mine)

And everyone knows what a leopard will do to a longhorn.

Yum! Yum! Dinnertime.

[Samanoske]

Originally Posted by sniffer

Perhaps you mean Schnappi ?

lol+

[Simon]

Originally Posted by wnuez

they might be on their way. have a look at: System->Library->CoreServices->Menu Extras->HomeSync.menu

Hmm, on my PowerBook's 10.4.1 I don't have /System/Library/CoreServices/Menu\ Extras

On what system did you find that?

[sandsl]

Originally Posted by Simon

Hmm, on my PowerBook's 10.4.1 I don't have /System/Library/CoreServices/Menu\ Extras

On what system did you find that?

/System/Library/CoreServices/Menu Extras/

[krove]

If you double click that menu extra and select "Sync Home now..." in the menu, you get this error message. This may actually be a feature of Mac OS X Server and require some extra software to enable this functionality.

[moodymonster]

Originally Posted by msuper69

(emphasis mine)

And everyone knows what a leopard will do to a longhorn.

Yum! Yum! Dinnertime.

ha ha

[Detrius]

Originally Posted by krove

If you double click that menu extra and select "Sync Home now..." in the menu, you get this error message. This may actually be a feature of Mac OS X Server and require some extra software to enable this functionality.

This is actually a feature of OS X Server, which I am using right now.

[Detrius]

Originally Posted by LeeG

Is there any chance, much like 'carbon' apps, that someone could write a compiler that would allow 1 binary to open both in OSX-Intel and Wintel?

It seems if someone could write a compiler that would allow this (as carbon did), that even MORE developers would write for mac - as they could output one program that ran on both.

Now I realize thats a lot harder than it seems, esp with different APIs for each OS, and also that apps would lose their "mac-ness" if they weren't developed specifically for macs, and to take advanatge of mac-only features, but it would be better than being told to reboot with windows and use the windows version....

L

Before Apple ever released the OS X Public Beta, they had the Cocoa APIs running properly on top of Windows. This was called the Yellow Box. In fact, it was released to developers. At some point, they decided not to use this.

Now, there are APIs like Qt and wxWidgets that allow you to write a program once and run it on multiple platforms. You do have to compile it for each platform. Also, this produces multiple binaries. However, the bottom line is you write the program once.

[Millennium]

Originally Posted by LeeG

Is there any chance, much like 'carbon' apps, that someone could write a compiler that would allow 1 binary to open both in OSX-Intel and Wintel?

Not without adding multiple extensions to Windows.

Although OSX-Intel and Wintel run code from the same architecture, that code is stored in different formats on both systems. Wintel uses a format called PE, whereas OSX uses a format called Mach-O (as a general-interest note, Linux uses a format called ELF). These formats are not compatible with each other, and although an operating system can be made to load code in multiple formats, it's not just a quick switchover. No code has been written to do this on either operating system.

Carbon works because both OS9 and OSX can use a format called PEF (which is different from PE). Both operating systems can load applications in this format, and so the format is not an issue.

However, even assuming that the binary-format issue is overcome, there are other obstacles. For one, both operating systems offer a different set of Application Programming Interfaces (APIs), which are pieces of code which developers use to make their apps work. These APIs are completely different, and although it is possible to rewrite the API of one OS so that another OS can use it -that's what Carbon is- this isn't an easy thing to do. Carbon took about two years to write for OSX 10.0, and even then it was horribly incomplete at that time.

Now, with all of this said, there's still the question of which OS should be extended and how. Should Windows be extended to include the OSX APIs, or should OSX be extended to include the Windows APIs? There are several reasons that I believe Windows should be extended, not the least of which is that it's not really true that no code has been written: many Mac APIs have been ported over, and exist in the form of QuickTime and WebObjects. This code is not perfect, and could not immediately be used as a complete OSX compatibility layer, but it would make a very good starting point. However, as long as it's an extension to Windows rather than being folded into the main OS, it will remain a niche player, because Windows users are (rightly) afraid of downloading software. Convincing Microsoft to do the work to fold in OSX APIs would be difficult, to put it mildly.

What about going the other way, then: porting the Windows stuff over to OSX? It wouldn't be without its advantages: Windows developers already know the Windows API, so they wouldn't have to learn a new one to develop for the Mac. However, if you wanted complete compatibility between the two APIs, then the result would be disastrous. The win32 API is riddled with fundamental flaws, mostly because of decisions made by Microsoft to trade stability and security for raw speed. Folding them into OSX would all but require those same decisions to take precedence, and although OSX would come out somewhat faster as a result it would have many of the same problems Windows does. That's a tradeoff which just isn't worth it. You could support a subset of the API, as Carbon does, but depending on what you keep and what you take out you'll alienate at least a few developers.

[yeyeogun]

Originally Posted by :dragonflypro:

10.5 Leopard.

Apple Changes Its Spots!

T

You stole my line. If you check Think Secret you will see a post where I used that comment to suggest that Nobody should understatement Apple because a Leopard will never change it's spots. I am willing to split the royalties with you if an advertisement comes out. Can you did that sucker!

[martman]

Originally Posted by iPhotoStuff

A big button on boot up called "Make OSX Function Properly" which the user can activate. Tiger seems to currently be missing that option.

Tiger works flawlessly in my world, which was a bit of a surprise since I run a gamut of test beds rtanging from a 400 Mhz iMac, to a 700 Mhz iBook, an 800 Mhz Cube, and a 1.66 Ghz Powerbook. The window redraws on the 400 Mhz iMac are in particular, suprisingly quick.

[MrToast]

Originally Posted by Krypton

If there were WINE for OS X, it would obviously be the killer app, whatever you happen to think.

It's called Darwine:

http://darwine.opendarwin.org

Cheers,

MrToast

[arglborps]

>It would also make developer's lazy, why spend money porting a Mac OS X application if you could use the Windows version through WINE?

Because nobody on the Mac would buy it? Don't forget Mac users are known to be buying more software on average than the average Windows user, you don't want to piss of good clients. Many professionals use the Mac because of its GUI etc. they want native apps that behave according to the Apple Interface guidelines. Many applications that were natively ported to the Mac have failed in the market, because the interface was too cobbled together.

WINE would only be a necessary evil for software that isn't available on the Mac, and where there's no native Mac application you could use otherwise. I for sure don't want to work with a Windows App with its completely fubar-ed interface, unless I absolutely have to. I know many Mac users feel the same.

[Okonomiyaki]

Originally Posted by midwinter

Once WINE is ported to this new Mac and Windows apps can run without emulating a different chip—and therefore at near-native (or native) speeds—why would MS or Adobe bother compiling a universal binary? Why not just bundle WINE with their software or say that Mac users need to have it installed?

I don't know... maybe because they want their software to sell better than Word 6 did?

[webraider]

Don't Get me wrong... I'm a big fan of Disk Utility... but I would think by now Apple would have made a very easy way to Format it's Optical media, or to do all the features of Disk utility with out having to launch an application. To me... the Writeable, re-writable media should be as easy to use as my zip disks are... Format them from the finder, Create a disk image from the finder... and pretty much be able to do everything you can in disk utility from the finder. It's not that hard to do and to me it makes sense. It almost feels like Writable optical media is an afterthought in mac OS X. Is it?????

Dare I say this is all easier to do in Windows????

[Swift]

My nomination for new feature is the integration of iChat with VoIP, with as many messengers as possible, and my most favorite feature, integration with Skype. Apple owners need a cross-platform, around the world, text, audio and video chatting program. I mention Skype because it's the one that's done correctly. This is one place that the elite feature -- the fantastic clarity of conversing with another G5 owner, or better, on fast broadband, is too exclusive a club. We need to have the best messaging service on the Mac. Skype is multi-platform, one. It has SkypeOut, two. The person you're talking to doesn't even have to have a computer, let alone a Mac. We should be able to converse with the wireless video phones of two years' time. They should be able to contact us. All computer platforms should be available. And yes, if you connect with another fast machine that has iChat for Windows or Linux or whatever, you can talk to them via videophone. Do it, Apple!

[LeeG]

Millennium-

Thanks for an excellent response - very well explained, I appreciate it - I assumed it was a monumental task, and I knew the API problem.

I also imagine most apps would end up looking like Word 6.0 - a windows app running on a mac. Part of the style and interface consistency is what I like about mac apps, and I am sure a hybrid would lose that polish.

Thanks again,

Lee

[starwxrwx]

Didn't Apple announce the next ver of OS X would be 10.6, not 10.5?