Honestly, I'm a bit surprised by the responses in this thread. The rumour might be true or false but that's not what I find surprising.
Are there technical problems regarding the rumour? There are, certainly. However, we developers know that technical problems can be solved and Apple and Microsoft certainly have the resources to do so.
Are there business issues? Is this an unexpected move? Certainly, it is. But Apple has a history of unexpected business moves and business innovation, as does Microsoft even if it is uncool to acknowledge that.
Even if the rumour is false (I don't know) I don't find it unbelievable at all, because (a) it makes sense (b) this move would be both typically "Apple" and typically "Microsoft", if you look at the history of these companies. Hell, Microsoft started as a programming tool company.
It does, as it works for a lot of people just fine, but it's not without it's quirks.
Some things that are better in XCode: filtering of options, error bubbles, keyboard mappings.
Some things that are better in MSVC: Debugging! Maybe because I had to use MSVC for the last 10 years, and I'm so used to it. XCode is also not bad, but would need more time adjusting to the keyboard shortcuts
Some things that are better in Eclipse: refactorings. Adding method parameters, changing their order, extracting methods, promoting or demoting classes, etc. At least when it comes to Java, Eclipse seems to have more options and a greater success rate than VS2008 with C#.
Debugging Java with Eclipse is about the same as debugging C# with VS in my experience, though it seems that VS's output console is slower than Eclipse's, and both are slower than a pure terminal.
I have to use both Eclipse and VS2k3/VS2k8/VS2k10 on a daily basis. Sure, eclipse has better refactorings. I would also argue that it's References/Declarations search features are better than the VS equivelents.
But that is the only two features where VS gets eclipsed. VS is an all-round better tool than Eclipse for everything else I need to do. Sure perhaps I can make Eclipse better by searching out some plugin written by some 3rd party, because the 3rd party support/community is better for Eclipse than VS. But from a daily use POV, atleast IMO, VS is simply leaps and bounds beyond Eclipse. VS also cost me $1000, Eclipse didn't cost me anything. So perhaps my view is skewed simply because I'm deluding myself because of my financial outlay, but I'd like to think not.
I don't know about keyboard mappings. I'd rather press F10 to step through code rather than apple key + Shift + P or whatever it is. I'd also rather press F5 to debug than apple key + enter. I also like that tab in visual studio tabs lines if you highlight them instead of replacing them with a tab like xcode does.
Actually I kind of put it wrong - I like Visual Studio debugging keys better, as I spent most of time debugging, but for typing (and since I've moved to emacs for my linux) I prefer XCode (it's not the same but closer).
I hate XCode debugging mappings (but I haven't debug enough on it) :)
They do, and I don't think that's going to change, because Microsoft has a team of developers whose job is to write the software that they use to write software.
This actually makes a lot of sense considering that right now if you want to develop for the iPhone you need to do so on a Mac. However, there are many developers who use PCs and who would also make good iPhone app developers, but legitimately have no means of doing so.
The technical issues are definitely not insurmountable. Mono, the free CLR implementation, had CLR apps running on iPhone years ago, and several commercial and prominent games were developed with Unity Framework (which is based on Mono). There was even a "MonoTouch" development profile for iP* for MonoDevelop, a popular Mono IDE.
So, not only are the technical issues solvable, but they'd been solved for years by the open-source clone of .NET.
Not all of them ... with both MonoTouch and Unity you need to have a Mac OS X running with XCode installed, otherwise you can't sign, package or test your application in the IPhone emulator.
If this announcement is about developing, packaging and testing iPhone/IPad apps with only a Windows workstation within Visual Studio, then that's certainly good news.
Not to mention that MonoTouch and Unity are legally both in a gray area right now ... because they are explicitly disallowed in the new SDK EULA and Apple hasn't mentioned their official position on them (AFAIK).
I want to clarify the legal issue, though. There is no (non-patent) contest on MonoTouch or Unity's right to existence; people are free to make iPhone runtimes, compilers, etc., as much as they want; the issue arises when developers deploy these techs, as Apple's new EULA is applicable only to developers. While the effect may be the same (no use), I think it's an important distinction.
So... Is this one reporter tossing his speculation on top of another reporter's speculation? And now leading to us speculating about his speculation??
I'm honestly amazed we are falling for this. It's exactly the same thing I see every holiday season at the grocery-store checkout: "Psychic reveals which star will die of a drug overdose this year, who is going to get married, and why a giant earthquake is going to flatten Toronto." It's a simple game that delights the rubes. Make a lot of guesses, take credit for the ones you got right last year, and hand wave the ones that didn't come true.
How is this any different?
p.s. Here's a suggestion: Create a poll asking HNers to vote on which wild-assed-guess is going to come true and keep all the talk of vapourware in one place!
At first blush, I dismissed it for violating 3.3.1, but now that I think of it, you're allowed to use Objective-C, C or C++. Visual Studio already has a C/C++ compiler so they don't even need to introduce a new language into the mix.
And since it looks like Apple is betting the company on the iPhone/iPad (as opposed to Mac OS X) there's no real downside to letting developers on non-OSX platforms build iPhone apps.
You would still need the $99 Developer License to load apps onto the device. And Apple gives Xcode away for free so they're not losing money on allowing competing development tools.
I still don't think this announcement is going to happen, but it's not the most outrageous thing that could happen.
If anything its more likely that Microsoft would reject such an idea than Apple.
> there's no real downside to letting developers on non-OSX platforms build iPhone apps.
To fully enjoy network effects you need to control the whole stack. This is the problem with this rumor: Jobs may be anything but he's not suicidal.
Allowing iTunes to run on Windows helped increase sales of iPods and music and preventing those who already had PCs running Windows to considering the competition.
Allowing VS2010 to target iPhone/iPod/iPad would increase the number of developers who could develop for the platform, something that I don't think Apple really needs.
If nothing else, Apple could raise the bar on apps a bit. This would only lower it.
If nothing else, Apple could raise the bar on apps a bit. This would only lower it.
On the contrary, if Apple had more apps to choose from, then it could afford to be more selective. The only thing is that they're already in a position to be more selective, and they're clearly not! (Perhaps too much work.)
That's the point. Apple doesn't need WinMo developers porting their apps to iPhone because a) they are terrible, b) there are not many of them and c) the important/interesting ones are already ported
At first blush, I dismissed it for violating 3.3.1, but now that I think of it, you're allowed to use Objective-C, C or C++. Visual Studio already has a C/C++ compiler so they don't even need to introduce a new language into the mix.
Section 3.3.1 also disallows using "compatibility layers" to access documented APIs. Since the Cocoa APIs are Objective-C, you wouldn't be able to write an app using only C++ without violating it (except perhaps in the case of an OpenGL game).
This doesn't even pass the giggle test from a technical perspective - iPad/iPhone is so massively tied to the Mac platform, how could you ever write any sort of meaningful integration with VS2010. You'd have to port all of the UIKit/CoreFoundation libraries over, you'd have to make a Windows version of the iPhone Simulator, you'd have to write a Objective-C compiler from scratch (remember, Microsoft would never ship GCC), that was compatible with GCC, the list goes on and on.
Now, maybe he's announcing a compatible version of Silverlight, or a way to compile SL applications to iPhone/iPad, which makes far more sense, and would be very compelling for developers. From one codebase, you could have a rich website, a desktop app, a WinPhone7 app, and an iPhone/iPad app, and you could write that app in C#, VB, Python or Ruby. That'd be awesome.
That was before GCC switched to GPLv3. However, the point is moot as someone else mentioned Clang, which I could see MS shipping. I did forget about SFU though, good point.
I completely degree with your assessment of the facts.
The debugger is GDB. This is not hard to integrate with on windows, as it already is integrated with on windows.
The compiler is gcc. Objective-C for gcc already exists on windows. Additionally, several people (Appcelerator, Adobe, Corona, Unity3d) have written machine code compliant compilers, I think Microsoft could likely produce one in a small number of weeks if they were being illogical about gcc.
Additionally, Microsoft has pretty much continually shipped gcc for the last 15 years or so depending on the products they were selling in the windows/unix integration sphere (and OS2 tools as well).
You don't need to port all the UIKit/CoreFoundation libraries over, just the simulated versions, which already run on x86. I'm pretty sure that could be done in a reasonable timespan, probably 3-4 months at the most. The rest are cross compiled for the device, and cross compiling is largely the same no matter what your host platform is.
There is 0 chance Silverlight is in any way involved as an implementation platform for general purpose use (but the iPhone simulator may have been ported using it). The entire point of the 3.1.1 language was to tie everyone to C, C++ and Objective-C.
Microsoft already ships cross-compilers (for ARM, for example) with VS. And Microsoft already ships dev tools for mobile platforms with VS (including mobile device emulators, etc.). Moreover, MonoTouch showed that this sort of thing (developing iPhone apps in C#, for example) is eminently possible.
From a technical perspective there are no serious roadblocks.
Now, whether or not this is actual news and not completely unsubstantiated rumor is an entirely different question.
You are uninformed about how the iPhone Simulator works. It is not, in fact, an emulator. It's simply a bunch of iPhone OS programs (SpringBoard, Safari, etc.) compiled against OSX and put in an iPhone-shaped screen. Your iPhone app is actually a userland Mac app compiled for x86.
So here's the problem: no amount of glue (short of shipping all of Snow Leopard for Windows) is going to get you an iPhone Simulator on the windows platform. It depends on OSX so deeply that you get weird behavior because of it (iPhone's accelerometer APIs are tied to your disk's accelerometer on the simulator, which is in the wrong orientation. Only your mac's built-in keyboard, not a USB keyboard, works in the simulator, because the iPhone uses really low-level keyboard APIs in OSX. etc.)
But they can't ship an emulator (or they would already have done so for OSX). Because except on very high-end systems, you're not going to emulate a 600-700mhz ARM plus their graphics chip. And obviously Silverlight & Co are firmly in the game development camp, where performance testing matters.
This makes perfect sense, at least from my anecdotal experience: there are tons of people who want to develop for the App Store but don't want to spend thousands to buy a new computer running OSX to do it. At this point, a developer who doesn't have an Apple laptop isn't likely to get one, so there are few hardware sales to lose, but a lot to win by letting Windows-based developers make apps for iP*s.
So the Windows coders who balk at a several hundred dollar 'Apple tax' are going to instead shell out several hundred dollars for a VS2010 license?
Or is the theory that Microsoft would let a feature like this wind up in Express when their own tools to develop for Windows Mobile are only available in vs2010 licenses that cost more than a macbook and app store license combined?
EDIT: Which isn't to say that it won't or couldn't be plausible. But if it does happen, it will be all about enterprises: The sorts of places that wouldn't blink at the cost of a Mac, but would balk at having to train, staff and support a completely different toolchain and environment.
The mac mini is more than a reasonable machine for development. I use it for gaming, design work in CS4 and development, when I don't feel like using my laptop.
That would mean that the 'tax' is only $500, or cheaper if you get it refurbished.
A lot of the developers who would be working on small apps for the iPhone/iPad would probably qualify as small businesses for Microsoft's BizSpark program, which gives them free access to lots of tools including Visual Studio. You qualify as long as your business is real, less than 3 years old, and earning less than $1mm in revenue. Once you exceed that, you should be able to afford the licenses.
Why would MS drive more consumers to the iphone only to retain developers? Consumers outnumber developers, what? 10000:1? Also, what signal would MS be sending to its developers about its own future platform.
More likely is some native office apps, bing search integration, and maybe silverlight news for MacOS.
I'm not saying you're wrong about this announcement, but honestly I see MS's strongest position right now to be with developers and in the enterprise.
It's not going to happen overnight, but I think we're seeing a pretty strong shift away from traditional computing for most consumer applications. MS really has no traction in that market (I suppose they could try to build off XBox), but they're still dominant in the business sector. As such I don't think it'd be a mistake for them to focus on developers and tools.
Worth pointing out that I don't see how this makes much sense from Apple's perspective either. Their march towards HTML5 is going surprisingly well, they aren't hurting for more apps or developers.
This'll be about Office 2011 on MacOSX, and maybe, maybe, some native office apps for the iphone.
I think even bing on iphone is out of the question.
I call BS - if something sounds to good to be true it usually is...
Far more likely that time is reserved for A) Bing over Google announcement in iPhone OS 4 or B) Office related announcements (maybe iPad?)
Given the investments Microsoft has made on the Mac the most agile team they have who could pull off something like this would be the Silverlight team who already has a .NET runtime (SL is a small .net runtime) running on OSX - one would imagine the jump to iPhone as rather trivial all things considered but there are two reasons this just doesnt make sense. Number 1 is the fact that since SL4 shipped the entire SL team focus has turned to WinPhone7 which uses SL as the app runtime. Second, given all the Apple pronouncements about being against 3rd party runtimes - I cant see them saying 'oh, Silverlight is OK'
What if the Silverlight development environment could target an Apple-developed and controlled runtime?
Apple would retain control, but gain access to an additional developer base as well as a Flash competitor with Apple platform control.
I'm not speaking for the credibility of the Register here. (I wish they'd die as much as Adobe would.) However, technically speaking, this is quite possible. It's also an interesting idea.
Add in a free SWF to Silverlight cross-compiler, and you have a very strong move against Adobe.
There is no such thing as an SL development environment, SL is a runtime and VS and Expression are the primary development environments. Semantics aside though, I just dont see it. Developing an SL runtime is not an easy task, it would take a large team at the very least a year to put one together. I could buy a situation where MSFT hands over the SL runtime source to Apple who then takes it over as part of some long term agreement to guarantee SL apps run on all Apple devices but that is about it.
There is no such thing as an SL development environment, SL is a runtime and VS and Expression are the primary development environments.
Sorry, should've used a plural.
Developing an SL runtime is not an easy task, it would take a large team at the very least a year to put one together.
Not true if there's cooperation from the MS development team. There would be compelling reasons for such cooperation. Flash is Silverlight's incumbent competition, after all.
I don't see how this is even remotely possible. How would you run/debug apps you compiled with VS? I highly doubt that Apple would port their platform-specific libraries and frameworks (among others, OpenGL.framework, Kernel.framework, IOKit.framework, IOSurface.framework) to Windows. What happens when you try to use mach ports in your code?
Without the ability to run the app you just compiled, this rumored feature is next to useless.
"How would you run/debug apps you compiled with VS?" --> Visual Studio could launch or connect to a virtualized iPhone device and let you debug there. That's the approach you use when developing Android apps with Eclipse ... Eclipse connects to a virtual emulated Android device.
I also think if this happens that Visual Studio would support Objective-C (and C/C++) instead of C#/Silverlight on the iPhone. Microsoft would prefer the C# approach but Apple won't allow it.
Virtualization could work, but Apple would have to build in support for it (versus the current approach of running x86 versions of the iPhone apps/frameworks).
Just wondering (I don't know much about OS X / iPhone development), why is it necessary to port it to x64? Surely OS X can execute 32 bit apps, just like Windows x64?
The point was highlighting how long it is taking them to do something which presumably is simpler than porting the whole (massive) program to a different OS.
Apple needs some positive publicity to help calm the ruckus over 3.3.1 changes, plus all the noise Adobe caused. They also would like to start calming any rumblings from the DOJ that are starting to tremble. This helps on all fronts.
Microsoft needs a positive win that says to the world, 'hey we are still relavent!', the mobile train might have left the station, but we're onboard!'. Also, this will certainly kick-up the VS2010 adoption rate, which will net MS a pretty penny too.
June 7th is right around the corner, so we'll know soon enough but this rings true to me.
I also think it is a win-win but for entirely different reasons.
Microsoft failed to get Visual Studio into the mobile app development world, simply because previously it only allowed developing for Windows Mobile which is, honestly, something nobody really cares about. With iPhone development support, they could get new markets for Visual Studio.
Apple wins because (I hope I won't offend anyone with that) Visual Studio is just simply vastly superior to XCode. It is a truly great IDE (as opposed to Adobe Flash) that could make OS X / iPhone development an awful lot easier.
Even if VS is better than Xcode (and I have no problem believing that it is), I don't see how that's a win for Apple. First off, Apple isn't in the game of admitting that their products suck, and they certainly wouldn't encourage people to go out and buy PCs.
Secondly, Apple doesn't make that much money off app store sales, they make by far the most money off of the hardware. This includes (though with decreasing relevance) mac hardware sales. This may not discourage all that many mac purchases, but will it really increase iPhone purchases? It isn't like Apple is having a hard time getting apps on the store as it is, they are already dominating in terms of volume. Not to mention, this would be reaching out to a world of developers that Apple has spent the last decade insulting as being incapable of producing quality products, exactly the kind of "crap" the "review" process and 3.3.1 are designed to "prevent".
The only argument in the back of my mind that strikes me as particularly persuasive is that Apple is gearing up way ahead of anyone's predicted schedule to simply stop making macs altogether, at which point they would need a solid alternative development platform.
The most important feature of the iPhone is the App Store. Many iPhone commercials and ads are simply showcases of apps in the App Store and this is not a coincidence. More (good quality) apps in the App Store means more iPhone, iPod Touch and iPad sales for Apple.
>> Secondly, Apple doesn't make that much money off app store sales, they make by far the most money off of the hardware.
Do you know this for a fact, or are you speculating? They don't break out app store sales in their 10-K, but consider quarterly hardware sales from Q4 2009:
- 3 million Macs
- 9 million iPhones
- 21 million iPods
In the same period, users downloaded 400 million apps, although it's unclear how many were paid. App sales are almost 100% profit, compared to the 20% net margins on hardware.
Of those 400 million apps, the best guesses suggest that fewer than 10% of downloads are paid. Estimates have put the values at about 10-20 cents per dollar app. The costs are fixed though, so they do make much higher profits on more expensive apps. But the app store ecosystem hasn't made that a huge factor.
Even under optimistic assumptions: 100% profit, average sale price of $5, 15% paid downloads, the numbers are not good:
400,000,000*0.15*5*0.3 = 90,000,000
That's nothing compared to the billions they are posting in profit each quarter.
The hardware margins are also almost certainly much higher than 20%, probably over 30% on a lot of the higher end hardware.
"simply because previously it only allowed developing for Windows Mobile which is, honestly, something nobody really cares about"
It's something nobody cares about now, but back in 2007/2008 (when the last version of VS was released) it was a sensible choice. The iPhone was new and there was no app store yet, there were no android headsets yet, and the palm pre didn't exist. Granted, things have changed enormously since then, and that's yet another knock against Visual Studio's slow release schedule.
Still relevant? I almost dismiss people out of hand who imply or make statements like this. Like it or not Microsoft is still the 800lb gorilla of the software world.
I find that this site makes more sense if you mentally do s/relevant/trendy or s/relevant/cool. It's annoying when people overload words with completely unrelated meanings, but there's not much you can do about it.
Apple needs some positive publicity to help calm the ruckus over 3.3.1 changes
I don't see this as an improvement. It just reinforces that their approvals and rejections are completely arbitrary, based on whatever they feel helps their business model.
But yes, it would definitely help Apple and Microsoft. It's in both their interests to kill Android and set up a duopoly where we get the "choice" of which closed platform to use.
I don't think this is real. If they wanted more developers and an influx of apps they would have partnered with Adobe. I don't think it makes sense to partner with Microsoft who competes on a hardware and software basis.
In the case that this is true, this would be far and away Apple's biggest-ever sacrificial move against the Mac for the sake of iPhone OS. Ending OS X's exclusivity as the dev kit for iPhone OS would send an extraordinarily profound signal about the company's priorities.
Question: How many Macs do you think they have sold to developers who (a) wanted to build an iPhone application but (b) didn't already own a Mac?
That's the market they are 'sacrificing'. I doubt it's significant.
update: Apple reports that they sold 3.05 million Macintosh® computers in Q42009 alone. I can't see them giving a damn about selling Macs to iPhone developers.
Humm I'm not sure what the actual rumor is.. VS 2010 for the Mac or just the ability to dev for the iPhone or the iPad using VS on Windows?
From the technical both would be possible but I guess there would be some effort required to do either of these:
- VS on a Mac: VS 2010 uses WPF for the gui. Therefore it would have to run on a Mac. I guess the .net platform probably already runs on a Mac (there is Silverlight for a Mac) and since WPF uses Direct-X they would need to hook it up with OpenGL. However since VS is huge this would take a lot of effort I guess, so rather unlikely.
- The second rumor, that the VS on Windows would be able to produce iPad/Phone apps seems easier to do from a technical perspective, since they could use Clang/LLVM. But I don't really see the point here: There are more then enough people developing for the iPhone/iPad and this would probably hurt Microsofts Windows Mobile plans.. as well as Apples Mac sales..
So it's a rumor - that's all probably complete BS. We will see..
I was thinking that what if they allowed a toned down version of SL on iphone, something without any of the localized storage libs. That would put it pretty much as a UI for any type of webservices (which is really no different than the web at that point). It would mitigate the ability for SL to perform features similar to apple apps (doesn't cannibalize app sales), but also would "get silverlight" onto the iphone. Something good for both MS (hey look at Silverilght, it went where flash couldnt!) and also for Apple ("hey we told you were not against other frameworks, look here we allow silverlight!").
This goes against everything they've been vehemently defending the past months. An extra layer, non-unique apps, etc. AND Microsoft is a much larger competitor than Adobe.
Don't get too excited, although I don't personally have authority to dismiss this, I have word that this is a completely FALSE rumor (sorry, I can't back that up-- at least do yourself the favor of being skeptical).
Iirc, first generation iPod did not work/sync on Windows, so what is so unbelievable about Apple having the iPhone SDK working on Windows with Visual Studio now?
Although personally I don't think this is going to happen. As a hardware company, i'm sure they make many extra sales from companies and developers purchasing for the sole purpose of targeting the iPhone/iPad/iTouch and Mac OS X users.
The App Store has enough crappy apps. I don't think opening the platform up to VS is going to bring in any more @atebits or the like. If you want to be serious about iPhone/iPad you will get a Mac.
I don't know if it's true or not but there's certainly smoke. Perhaps aimed at taking the thorn out of the section 3.3.1 changes relating to third-party tools?
I actually understand Steve's point about cross-platform GUIs being the lowest common denominator (just look at Java Swing apps) but it's disappointing that change also hurt MonoTouch, which was a 1:1 iPhone API mapping and not an intermediate layer like the Flash compiler was/is.
Apple should have a certification program for such 1:1 mapping APIs. That would also take the thorn out of section 3.3.1 while allowing them to retain platform control.
>I actually understand Steve's point about cross-platform GUIs being the lowest common denominator
Sure. iTunes on Windows.
>I don't know if it's true or not but there's certainly smoke
All of the smoke emanating from a single nobody making a completely unsupported claim. It's actually a bit extraordinary how little is necessary when it comes to Apple.
I always wondered why iTunes for Windows is Carbon based. NeXT had a full OPENSTEP stack capable of running under NT (I did it) and that is what Cocoa is based on.
Are there technical problems regarding the rumour? There are, certainly. However, we developers know that technical problems can be solved and Apple and Microsoft certainly have the resources to do so.
Are there business issues? Is this an unexpected move? Certainly, it is. But Apple has a history of unexpected business moves and business innovation, as does Microsoft even if it is uncool to acknowledge that.
Even if the rumour is false (I don't know) I don't find it unbelievable at all, because (a) it makes sense (b) this move would be both typically "Apple" and typically "Microsoft", if you look at the history of these companies. Hell, Microsoft started as a programming tool company.