Dharma: Cocoa on Windows... Again?

[BillyShears]

Let's look at developers, since what benefits developers benefits Apple. (DEVELOPERS! DEVEOPERS! DEVELOPERS!!) This benefits Apple if more developers make apps in Xcode, thus being compatible with OS X and Windows.

But why would the developers do that? I don't think they would.

First, they're (most likely) not trained in Xcode or Cocoa.

Second, the application (most likely) wouldn't "fit". There are differences in OS X interfaces and Windows interfaces. Such as: persistent menu bar, application name as a menu bar item, location of "OK" and "Cancel" buttons (actually Apple specifies not to use the word "OK"), etc.

Third, as someone mentioned, the frameworks would need to be ported. But as a Windows user I'd expect an application to get my contacts from Outlook, not from Address Book.

So for a developer, you need to learn new things, re-arrange your interface, and use different frameworks. That's not to say it's impossible (see: Firefox), it's just not "clicking a checkbox".

Even Firefox doesn't (last I checked, not 1.5) look & feel right on OS X.

I guess it would be somewhat beneficial for a developer. But without significant work, your app wouldn't look or behave right for all the Windows users (~95% of users?). As opposed to developing for Windows and recommending WINE for OS X (and Linux) users, which would keep it looking right for the Windows users (majority) while letting the OS X and Linux users suck it up. Or am I missing something obvious?

Also, what apps would be converted? Are they trying to move Windows apps to OS X or OS X apps to Windows?

From Windows -> OS X... Brand new apps might go, if the developers have time to learn Xcode & Cocoa. There would need to be some EASY way to get medium sized apps to move over. As for large or custom apps? I doubt it. It would be a big effort, I would think, to move from Visual Studio .NET to Xcode. For a big project, I don't see it happening.

From OS X -> Windows... this is the way most apps would go. Since they're already in Xcode & Cocoa, I can see it being a lot easier (and therefore more common) to make an OS X app work in Windows and OS X.

But we want Windows apps on OS X, not the other way around! Windows developers will say "use WINE" and OS X developers will offer EXEs for Windows.

Now, as for Apple. Where is the benefit?

If all OS X apps run on Windows, why would anyone buy a Mac?

Releasing Safari for Windows would certainly be useful. It would let web developers test without buying a Mac. For the most part, applications where standards are at play (HTML, OpenDocument, etc.) I could see it as beneficial for Apple to have a presence on Windows. More Apple marketshare means more of their standards supported, which means more people can move from Windows to OS X.

It would be a good idea to keep it in mind. I can see such a development environment being useful at some point in the future.


Maybe I'm being pessimistic, but I don't see much benefit for Apple or for developers.

What I can see happening is this: Apple ports Cocoa to Windows, but keeps the Xcode builds that can make Windows binaries INTERNAL. Then they can selectively release apps they need on Windows (Safari, Pages, apps that involve standards & marketshare).

That's the way I would see it going down, anyway.

 

[adrianm]

did you read it?

Guys, this is 100% fake. The guy put the project name as Dharma, and signed the email "John Locke, somewhere near Hawaii".

Watch LOST on ABC, and tell me it isn't fake.

S/he did say in the post why s/he was using a fake name. Duh.

 

[Morn]

Yeah, Safari on Windows? Why lets put final cut pro and logic on windows also, and lets make a dock for windows, a new finder based shell to replace explorer.exe. Why lets remove any reason for people to get a mac.

 

[gekko513]

BillyShears:
I disagree. Learning Cocoa and Objective C will be a much more attractive alternative to learn for developers if they know they can use it to program stuff for Windows, too, even if it's not ideal.

Also, it wouldn't be so bad to develop an application using Mac OS X's user interface guidelines and make it available for Windows, too. Lots of applications for Windows don't care about the Windows user interface guidelines anyway, so there's nothing new there.

 

[GregA]

to add my own comments about this... what about interfaces? would an Xcode user be forced to use the Aqua interface on Windows? because that would really suck, if I was developing for Windows I wouldn't want my app looking like it was emulated, just like I don't want any app with Windows interface running on my Mac.

Openstep was designed to take on the interface of the OS it was run on. To do it properly, you had to manually say where some menu items went (as different OSes do thing differently), but other than that it just took on the correct look AND feel (I read this from some programmers years back).

This was related to the early work on multiple "skins" for OSX... the code itself was independent of the look. I don't think Carbon (or classic) is quite as flexible.

 

[BillyShears]

BillyShears:
I disagree. Learning Cocoa and Objective C will be a much more attractive alternative to learn for developers if they know they can use it to program stuff for Windows, too, even if it's not ideal.

Also, it wouldn't be so bad to develop an application using Mac OS X's user interface guidelines and make it available for Windows, too. Lots of applications for Windows don't care about the Windows user interface guidelines anyway, so there's nothing new there.

Even so, there's a second part to my argument that says that this will encourage developers to move from OS X to OS X and Windows, not the other way around.

The interface problems can be resolved, it just takes effort. The bigger problem is all the frameworks. It seems to me they would just be building OS X on top of Windows if they implement all the frameworks.

And finally, if you're going to make a program run suboptimally, you're going to make it run suboptimally on OS X, not Windows. That means WINE, not Cocoa for Windows. I guess if they could make it look and feel like a Windows app, have integration into other Windows apps, then it might be cool.

But then why buy a Mac at all if all the apps are available for Windows? This is different from OS/2 supporting Win16 apps, I think. It could have the opposite effect: If all OS X apps are available on Windows, nobody buys a Mac. What they would be looking to do is get new apps & old Windows apps ported to Cocoa. I think they need to do more than just offer Xcode for Windows in order to do that. I think they need to offer an easy transition from Visual Studio.

 

[Morn]

Windows users tolorate slight interface guidelines errors out of habit. But if you start giving them a totally different interface design for a different GUI, I think you'll start to hear windows users saying, hey this sucks, works nothing like my other programs. I think itunes suffers from this a bit on the windows platform.

 

[GregA]

But then why buy a Mac at all if all the apps are available for Windows? This is different from OS/2 supporting Win16 apps, I think. It could have the opposite effect: If all OS X apps are available on Windows, nobody buys a Mac. What they would be looking to do is get new apps & old Windows apps ported to Cocoa. I think they need to do more than just offer Xcode for Windows in order to do that. I think they need to offer an easy transition from Visual Studio.

Aren't you arguing the opposite to most people? I've never heard anyway say "I'm getting a Mac for the 3rd-party applications"... that's usually touted as the reason to get Windows.

I do agree that the key here would be to make an easy transition from Visual Studio.

 

[eSnow]

Another skunk-works project

I don't see this as a piece of software that Apple will release soon. More likely, it is just another under-cover project that aims to keep Cocoa portable. Apple is just hedging their bets here, imho.

The precursor to Cocoa, the NextStep/OpenStep API was available on Windows, even after Apple took over NeXT. Back then it was named "YellowBox" and soon afterwards pulled much in the same way OS X/x86 was pulled earlier.

Seeing that Apple has secretly kept the x86 build train of OS X alive for 5 years, I hardly believe they just let the YellowBox code rot - more likely they have a small team of dedicated engineers that work to keep it in sync with the OS X build. Since a lot of Cocoa is fairly high-level code, this might be easier as one would imagine.

Porting Quartz and its Composer is much more work, of course, but if I am not mistaken, Apple has done this already - for QuickTime and iTunes. Seeing Quartz Extreme or even Quartz 2D is, however, extremely unlikely.

If I am right, Safari on Win is nothing more than a proofing app - something you build to show your port actually works. Apple will most likely have ported more apps like iCal or iPhoto but I would not expect any of those to be released very soon - unless they see themselves in a situation where it makes sense. Like when they released OS X/86 as IBM stopped caring about aggressively developing further PPC970xx variants.

 

[gekko513]

Windows users tolorate slight interface guidelines errors out of habit. But if you start giving them a totally different interface design for a different GUI, I think you'll start to hear windows users saying, hey this sucks, works nothing like my other programs. I think itunes suffers from this a bit on the windows platform.

Which media player on Windows doesn't have a totally different interface? Windows Media Player works nothing like any other programs, Winamp totally different interface, Real Player totally different interface.

 

[isgoed]

Let's look at developers, since what benefits developers benefits Apple. (DEVELOPERS! DEVEOPERS! DEVELOPERS!!) This benefits Apple if more developers make apps in Xcode, thus being compatible with OS X and Windows.

But why would the developers do that? I don't think they would.

[…] I guess it would be somewhat beneficial for a developer. But without significant work, your app wouldn't look or behave right for all the Windows users (~95% of users?). As opposed to developing for Windows and recommending WINE for OS X (and Linux) users, which would keep it looking right for the Windows users (majority) while letting the OS X and Linux users suck it up. Or am I missing something obvious?

What you are missing is that this is the situation as it is now. Wine and Virtual PC will be rock solid on OSX86. If apple doesn't watch out they will suffer the same fate as OS2 and nextstep. Apple is trying to change this by trying to make xcode the default developing environment. And then the same arguments holds as you were saying only the other way around: "Why develop on the windows platform if xcode is all you need". But the key is indeed (as you said) to convince the developers. For large companies such a transition is probably very hard, but there are a lot of mac-enthousiast developers and multi-platform software projects that already make great mac products. If they all move to xcode development only they can have a great impact on what the defacto-standard will be.

Also, what apps would be converted? Are they trying to move Windows apps to OS X or OS X apps to Windows?

• New apps made by developers who are familiar with cacao.
• Apps which are currently open source. This will have a humengeous impact since most of the time these apps lack a user-friendly-interface/installation (think Linux)

Windows -> OS X or OS X -> Windows isn't really the issue, because on mac and windows you can get a quite similar range of apps. The pro-applications (Adobe apps, 3D apps) are already developed for both OS'es. There are indeed a few pro-apps that are windows only (Engineering, CAD, business and such). They will not be transfered, but apple hopes that if this strategy works that eventually xcode gets enough momentum to also get those apps being converted. The one category of apps that still will remain on windows is the games. They require a lot more then just a cross-platform user interface API (cacao). But I don't think Apple will care. Consoles will probably dominate games in the coming 5 years and Applemac''s will be dual bootable to play a game.

Now, as for Apple. Where is the benefit?

If all OS X apps run on Windows, why would anyone buy a Mac?

Apple is indeed taking a big risk. Will there be enough differences to justify buying an applemac.

• On the one side: if they do nothing, they do keep OSX exlusive, but WINE might take over which might spell disaster.
• On the other side: if they release xcode 'for windows' they might lose the exclusiveness of OSX, but can attract more developers.

The former might be undesirable since currently most apps are windows only and for apple it would be BAD strategy to rely on WINE or virtual PC and it will keep the balance still at a 95% for windows. The attractiveness of the latter option becomes apparent when I answer your second question. Why do people buy macs now even if 95% of the software is developed for windows? Because a mac is the computer that makes life/work easier, it is reliable and it is ok to have it in your living room. It might be a kind of X-factor and only by retaining that specialness, this strategy will succeed. Right now apple has momentum, but they need to stay on top of the game for this to work out (not like windows that copied the mac and became dominant).

Releasing Safari for Windows would certainly be useful.

Hardly, I think. Ok those web designers have a somewhat convenient way to do testing, but that's really insignificant and besides the safari render engine is already available for windows so making Safari-win32 is just a proof of concept.

Maybe I'm being pessimistic, but I don't see much benefit for Apple or for developers.

There is indeed not much benefit for developers (your pessimism is justified), but still they are the key to apple's growth. Apple would thus be wise as to make the transition as painless as possible and introducing the benefits of truly universal binaries.

What I can see happening is this: Apple ports Cocoa to Windows, but keeps the Xcode builds that can make Windows binaries INTERNAL. Then they can selectively release apps they need on Windows (Safari, Pages, apps that involve standards & marketshare).

That's the way I would see it going down, anyway.

Hmmm, I don't agree actually. Why try to compete on the windows software market, when what you actually want is people to move over on to OSX and apple-PC's.

And might I add that I can see that this is technologically plausible. Apple says that on OSX86 version of an cacao app just takes a recompile and with nextstep already being proven concepts and the MacOS-folder within app's (so all you need to do is add an Windows-folder) the architecture is already there.

I have doubts about the credibility of the source, but regardless of wether the source is valid, I can see this happening.

 

[meat-eater]

This make me so happy!

I really, really hope this to be true. It would mean I can finally use cocoa for our projects even for the windows ones. To celebrate, i wrote a celebratory post on my blog: Cocoa on Windows

 

[wrldwzrd89]

<snip>
ps. I wonder whether you could run a universal application on OSX-Intel, then switch to Windows and open the same universal application.

Unless Universal Binaries are designed to determine the correct binary to use on first run and delete the other ones (which would save space once the application is installed, but would sacrifice portability, a reasonable tradeoff), this should work. In fact, if Apple is smart they'll implement property list support on XCode for Windows so that applications compiled in XCode are truly portable - so portable that you can simply drag and drop the application package and its corresponding settings file to another computer (PPC Mac, x86 Mac, x86 Windows, others?) and it won't even notice anything changed!

 

[Metatron]

So, apple is scared it will turn into nothing more than a harware manufature competing with dell, gateway, sony, etc, etc.

I can see why yellow box might still exisit.

 

[SiliconAddict]

Begin these API wars have...
-Yoda as a programming Guru​

Microsoft is releasing Windows Presentation Foundation to OS X...Looks as if there is a possibility of Apple countering.

 

[DavidLeblond]

I got my hands on an early copy of Safari for Windows. It seems pretty rock solid, except every 108 minutes it crashes. Something about me not hitting a button? I don't get it.



(I'm kidding of course, but I don't believe that email for a minute)

 

[ctachme]

Aren't you arguing the opposite to most people? I've never heard anyway say "I'm getting a Mac for the 3rd-party applications"... that's usually touted as the reason to get Windows.

I do agree that the key here would be to make an easy transition from Visual Studio.

Umm... the Apps ARE the reason I get a Mac. I get a Mac because it has Safari, and iChat and iMovie and all of those great Apple Apps. If those apps were on windows, what modivation do I have to get a Mac? And for evey app ported to windows, that's one less reasion to get a Mac.

Meaning I think the porting of apple apps is a Bad Thing.

However, the frameworks are fine. Let 3rd party developers get their programs made. Just don't get rid of anyone's reason to get a mac.

 

[GeeYouEye]

So, apple is scared it will turn into nothing more than a harware manufature competing with dell, gateway, sony, etc, etc.

I can see why yellow box might still exisit.

Scared? Hardly. Apple IS a hardware manufacturer (or at least designer and retailer now, since most everything is actually manufactured in Taiwan). Dell is Apple's main competitor, and has been since it took over that role from HP and IBM before it, not Microsoft. The OS is just a goodie to get people buy the hardware, which comes at a something of a price premium because Apple doesn't have the volume of Dell, and Apple wants a greater profit margin.

Personally, I don't see this happening at all.

Operating Systems, or at least the actual core OS, excluding anything fancy like a GUI, are going open-source. There are very few things you can do for an OS to compete with open-source software. Filesystems are one thing, but HFS+(Journaling, +xattrs) is not all THAT stellar, and Sun's ZFS looks to be the end-all be-all of filesystems. And it's open-source. So by default, Apple can't win with a proprietary OS. The best they can do is try to influence which OSS OS is in greatest use. And they're doing exactly that, with Darwin (though it seems they're not doing so well with that...). And that's not to say that there's been no innovation in the OS layer; thread-safe malloc() in libc is one thing (Other Unicies (Linux esp.) don't do this), even if it hurts performance a bit, it makes development easier.

Above the actual OS you have Applications and Application technologies. This is where Apple, or anyone else, can be proprietary and innovative, and still make money. Quartz, QuickTime, Cocoa, Carbon, Finder, Spotlight, Dashboard, Automator, iLife... these are unequalled in the computing world (they may have substitutes, but not equals... well, Finder not so much maybe). Apple bundles them with the OS (and calls them part of the OS, though Darwin can exist without them) because they don't need to make money on them; they're used as more goodies to stimulate hardware sales. But what goodies they are! Would anyone buy Apple hardware if it wasn't for what Apple calls OS X (despite the fact that the only real OS in OS X is Darwin, which can be downloaded freely)? Nope. So these are Apple's crown jewels. If the source code was ever to be released, or any of the technologies ported to other platforms (note that the ONLY QuickTime applications on Windows are QT Player and iTunes, whereas there's dozens or hundreds of apps that use the QuickTime API or QTKit on the Mac), it would be the end of Apple as we know it. They would be forced out of the hardware business, because the hardware is expensive and the software is readily available on other platforms. (This is also why Apple is so paranoid about people pirating Tiger x86 that they're putting TPM chips in their computers... frankly, I'd rather they stuck with PowerPC because "trusted computing" scares the crap out of me and is going to make me extremely wary of buying a Macintel) If Apple survived the transition to a software/consumer electronics company, they'd be directly competing with Microsoft, which is a battle they've long since acknowledged they can't win (though I suppose it's possible they might survive awhile, and maybe take away some market share from MS).

So I'm dismissing this rumor, interesting as it would be, as extremely unlikely.

 

[artifex]

I hardly ever use Safari, because it's so bogged down when running javascript, etc. I copied my Firefox profile over from my XP box as soon as I was up and running, and that's what I depend on, on a daily basis. I see no use for Safari on XP other than as an intellectual exercise, because nobody's shown me anything it can do better. Well, except this latest FF seems to crash

 

[Cochrane]

I do not believe in this rumor at all, but still, it would be interesting to see it happen. I'm developing a game for Mac OS X with Cocoa and Objective C. All my friends urge me to create a Windows version as well, so I'm currently looking into GNUstep, but it just sucks. I was not able to compile their Xcode equivalent yet and start working, so I really hope there'll be an Apple solution to this.

I wonder whether they should port Xcode. Not doing this would mean "Buy a Mac if you want to use this!", which might be good for Apple. But it would make debugging on Windows much, much harder.

 

[isgoed]

Personally, I don't see this happening at all.

[…] Above the actual OS you have Applications and Application technologies. This is where Apple, or anyone else, can be proprietary and innovative, and still make money. Quartz, QuickTime, Cocoa, Carbon, Finder, Spotlight, Dashboard, Automator, iLife... these are unequalled in the computing world (they may have substitutes, but not equals... well, Finder not so much maybe). Apple bundles them with the OS (and calls them part of the OS, though Darwin can exist without them) because they don't need to make money on them; they're used as more goodies to stimulate hardware sales. But what goodies they are! Would anyone buy Apple hardware if it wasn't for what Apple calls OS X (despite the fact that the only real OS in OS X is Darwin, which can be downloaded freely)? Nope. So these are Apple's crown jewels. If the source code was ever to be released, or any of the technologies ported to other platforms (note that the ONLY QuickTime applications on Windows are QT Player and iTunes, whereas there's dozens or hundreds of apps that use the QuickTime API or QTKit on the Mac), it would be the end of Apple as we know it. They would be forced out of the hardware business, because the hardware is expensive and the software is readily available on other platforms. (This is also why Apple is so paranoid about people pirating Tiger x86 that they're putting TPM chips in their computers... frankly, I'd rather they stuck with PowerPC because "trusted computing" scares the crap out of me and is going to make me extremely wary of buying a Macintel) If Apple survived the transition to a software/consumer electronics company, they'd be directly competing with Microsoft, which is a battle they've long since acknowledged they can't win (though I suppose it's possible they might survive awhile, and maybe take away some market share from MS).

So I'm dismissing this rumor, interesting as it would be, as extremely unlikely.

This is not about apple porting its i-apps to windows. It is about attracting more 3rd pary developers to take on the battle with the windows environment. And in this battle developers are your soldiers.

iApps will remain OSX only.

and you can make programs that integrate quicktime also on windows.

 

[cubist]

GeeYouEye is right, up to a point... That point is this: If Apple thinks it can keep making money as a boutique hardware supplier, they're dead. PC hardware has become a commodity, and any company betting otherwise, except for specialized applications, will not see the end of the decade.

As GeeYouEye rightly points out, operating system software, and much application software, is becoming a commodity also. That's why making Cocoa cross-platform is an amazingly perceptive thing to do - and if Apple isn't doing it, someone else should. It will be possible to have "that which is a Mac" running on a variety of operating systems, by having an "application platform" above the OS.

A question, though: How will Apple make any money in such a brave new world? By selling iPods, or developer support? It hardly seems likely. They will have to charge for something that they're currently giving away, I should think.

 

[isgoed]

As GeeYouEye rightly points out, operating system software, and much application software, is becoming a commodity also. That's why making Cocoa cross-platform is an amazingly perceptive thing to do - and if Apple isn't doing it, someone else should.

They already have: Real Basic

A question, though: How will Apple make any money in such a brave new world? By selling iPods, or developer support? It hardly seems likely. They will have to charge for something that they're currently giving away, I should think.

Apple will remain a hardware company. (Big) If hardware goes bankrupt they have to be like microsoft.

 

[Le Big Mac]

my thoughts exactly. Maybe Locke got bored of inputting 4 8 15 16 23 and 42 and tried to port Safari to Windows instead.

BTW, isn't that an Apple II (or ][) that they're using?

 

[skellener]

I wonder whether they should port Xcode.

The technology of putting "Universal Binaries" onto Windows and having apps that will run has existed for quite some time. Remember how Steve said they have had Mac OS X secretly running on Intel all this time? Well guess what, all of this work has already been done as Openstep. An Openstep app already could run on top of Windows. It wasn't even kept secret. It was available through Web Objects back in the day. Once installed, anything that had been compiled for it ran just fine. I remember seeing the demo by Steve on Openstep Day August 14th 1995. Web Objects with all of the Windows binaries was available through Apple well after the NeXT and Apple merger. It exists, it works.

So the question isn't whether it will work. It already does. The question is whether they will release it or not. I'd rather see it die myself. Let Windows users live in their crappy Windows world with sub-par applications.