The fate of third party application development for the iPhone is one of a few objects of speculation for Apple's Worldwide Developer Conference next week. What is likely in view of Apple's existing stance on the iPhone’s platform? Here's a look, leading into some other predictions for WWDC 2007.
The Other Big Event.
After Apple announced plans to pull out of IDG's summer Macworld Expo in 2002--the year the trade show owners decided to move it from New York to Boston--the company has turned its own WWDC into the second largest Apple event of the year next to the January Macworld Expo.
Both events are held in San Francisco at Moscone Center. Macworld Expo is held in the cavernous underground Moscone North and South conference halls, while WWDC is held in the new Moscone West expansion built just a few years ago. Moscone West offers a number of rooms for concurrent sessions.
At WWDC, Apple presents separate tracks of sessions suited to developers specializing in Mac development, QuickTime and Pro Apps’ content and media, IT management, and graphics and imaging. Apple also presents design awards, a ‘stump the experts’ panel, hands-on labs, and of course, the opening keynote address that unveils new public information with the dramatic spectacle of Steve Jobs and perhaps a new version of Keynote.
Previous years’ WWDC keynotes announced new information about Mac OS X, and occasionally introduced new hardware:
-
•2002: previewed Mac OS X 10.2 Jaguar; Jobs staged the burial of Mac OS 9.
-
•2003: delivered the Power Mac G5 surprise, Mac OS X 10.3 Panther, Safari, Xcode, iChat AV and iSight.
-
•2004: was all about new features in Mac OS X 10.4 Tiger.
-
•2005: announced plans to move the Mac to Intel within a year.
-
•2006: previewed Leopard and released the new Intel Mac Pro.
Compare Apple’s Macworld Expo announcements over the same time period, where the majority of Apple’s new consumer hardware and software is announced:
What's going to happen this year?
Despite the interest in third party development for the iPhone, a software development kit for it will almost certainly not be released at this years WWDC. While some have called for or predicted the release of an iPhone SDK, the only session related to the iPhone is one about developing websites that work well on its browser.
If Apple wanted to create developer interest in the iPhone, it wouldn't unleash it as an unbilled surprise at WWDC. Developers hate surprises. Apple spent most of the 90s slowly learning that.
Consumers love surprises, and flock to hear about the latest gadgets. Of course, developers like gadget surprises, too. However, developers mainly want to learn about what Apple has planned for the future, how they can take advantage of the latest technologies, and how new products and initiatives might make their jobs easier or open new prospects.
There is no doubt many developers are very interested in working on the iPhone. It therefore would be both foolish and a cruel trick for Apple to schedule nothing about the iPhone at its developer conference, then hold a surprise crash course in iPhone development along with a full SDK.
Imagine how many mobile developers would be upset that Apple presented just what they wanted to hear at a conference billed as being only all about Leopard on the Mac.
Should they be expected to pay $1500 to attend, just as a gamble that some iPhone bread crumbs might slip?
Grab the Dashboard.
While a full iPhone SDK appears to be almost certainly out of the question, it is interesting that Apple is holding that session on creating iPhone friendly websites. Doesn't the iPhone have a “real” browser, not some WAP or toy Internet that requires specialized mobile support?
Either Apple’s marketing is overreaching, or it is very likely that the company will be showing developers how to adjust standard Mac OS X Dashboard widgets to work on the iPhone.
Widgets are just a bit of HTML and JavaScript, essentially a mini web page. While some widgets can launch other applications or tie into more sophisticated functions, the majority of what they do is fantastically easy to port to other platforms, particularly ones that already run Safari’s WebKit, as the iPhone does.
That means it is very likely that the iPhone will gain support for third party apps in the form of widgets. That would expose a lot of functionality without raising the issues of security, stability, and malware which immediately opening up the entire platform might risk.
The majority of ideas users have written me about could easily be delivered by third parties as widgets:
-
•Converters and Calculators, whether for recipes, subnet masks, or languages
-
•RSS Newsreaders
-
•SSH remote command line access
-
•sFTP file transfers
-
•Mobile mini-games
-
•Flight, Transit and Package Tracking
-
•All manner of other web service front ends
This would require very little significant work for Apple, blaze open a wide range of functionality for iPhone users, and unleash a lot of opportunity for developers, all without waking any of the gremlins Jobs worried about.
Specialized widgets could be installed by the user via iTunes or even sold through the iTunes store, creating a business model for widgets that doesn't currently exist.
How Now, Cocoa Cow?
Widgets for the iPhone might make more sense than they do now on the Mac OS X desktop. Still, there are lots of advanced applications that demand a more than the Dashboard environment supports.
There is another level of functionality available in Safari, as WebKit itself supports a richer environment for more involved web applications. This kind of support, which is also likely be outlined at WWDC, would enable such applications as:
-
•Google Docs and Spreadsheets
-
•Microsoft's full Outlook Web for Exchange Server
-
•A web client for IBM Notes
-
•Various other rich AJAX applications
Still, there are remaining reasons why developers would like a crack at the full Cocoa, Apple's internal frameworks for Mac OS X applications. First, many Mac apps that would make sense to port to the iPhone are already written in Cocoa, so migrating them to the iPhone using the same Xcode tools would be far easier than building a web client or web version that can work on the iPhone.
At the same time, a lot of Mac applications are based primarily in Carbon, Java, or some mix of the three that would obstruct a simple port to the iPhone. Apple's own iTunes appears to retain its Carbon origins, and we know most everything from Microsoft and Adobe remains frozen in carbonite from the Classic Mac System 7 era.
As I noted earlier, there is great no emergency to deliver major new applications for the iPhone, unlike other new platforms which cry out for apps from day one. The iPhone already delivers at least $449 worth of the top titles of Windows Mobile software, fresh out of the box. What else might it need? Send in your suggestions!
Coming This Summer: Flip-Flops and BREW.
Writing on the subject of the iPhone and third party software for ExtremeTech, Bryan Gardiner says Jobs was “flip-flopping,” based on the initial comments made about the iPhone’s closed platform, tempered with Jobs more recent assurances at All Things Digital that in “an important tradeoff between security and openness [...] We want both,” and “I think everyone can get what they want.”
Flip-flopping sounds a lot like waffling. What a liberal Jobs is! He responds to panic in the market by offering a broader explanation of what's going on behind the scenes! How outrageous.
In any event, Gardiner offers a series of possible explanations as to how Jobs might deliver security and openness at the same time, just the kind of thing those middle-of-the-road Blue State American centrists are always trying to do.
iTunes vs BREW vs Open Shareware.
Gardiner makes a curious comparison of Apple's iTunes software distribution with Qualcomm's BREW. It should have been noted that BREW is just a noxious clone of mobile Java, except that instead of being somewhat problematic and mostly open, BREW is mostly problematic and not open in any respect at all.
BREW was conjured up mainly to serve the needs of Verizon, allowing the mobile service provider to rent software. Guess how I feel about BREW? It’s similar to how I feel about Verizon, if that isn’t yet obvious.
The differences between the business model of BREW and iTunes can best be made by comparing what they actually deliver to the consumers who use them.
Verizon rents a simple BREW version of the card game Texas Holdem for $3 per month. Apple offers the same game, with less imagination required, for the iPod in iTunes for $5 total; it does not vanish if you stop paying.
Compare both models to the “open” Palm software market, where the same game costs $20. Now compare real screen shots and decide what you'd like to pay to play cards on your mobile for a year:
-
-
-
•$20 for Palm shareware in an open software market (not open source)
-
•$5 for a nice looking game from iTunes
-
•$36 in rent for Verizon’s simplistic BREW applet
Gardiner doesn't point out those real differences, instead simply jumping to the conclusion that with an iTunes software model, “unlike Palm, Windows Mobile, and Symbian-based phones, the iPhone will not enjoy the same robust developer ecosystem for quite some time.”
Judging from the pathetic state of pricey mobile software on all those platforms, I would certainly hope not! But since Apple covers the bases of $449 worth of top Windows Mobile software, perhaps iPhone users won't notice.
Who is it availing themselves of this rich buffet of existing mobile software anyway? The problem with making up stuff that isn’t true is that you end up misleading people! Or perhaps that was the point?
LLVM: Drop it Like You’re Wired.
Gardiner also mentions the low hanging fruit of widgets I noted above. But the reason reader Robert Boylin forwarded me a link to Gardiner's article was principally due to comments about Low Level Virtual Machine. LLVM is an open source project Apple has sponsored for some time now, and appears to be weaving into Xcode.
LLVM is a fresh and modern approach to developing the tools to develop software. It assembles together code compiler components using a modular new design. This opens up the number and range of programming languages that can be added in the future, and makes it easier to plug in specialized code optimizers that benefit any supported language, even scripting languages like Ruby, Perl, and Python.
This new code compiler toolset promises to help streamline and optimize application development in Leopard, a task that also applies to the iPhone because it also runs Leopard. However, LLVM does not immediately solve the gremlins Apple has on its radar as it plans how to deliver a new generation of mobile software development.
Wherefore ARM Thou, Apple?
Gardiner determined that LLVM was the appropriate buzzword to associate in connection with the iPhone because of comments made in an LLVM presentation regarding support for the ARM processor architecture. “Why would Apple be concerned with ARM support, a platform that was certainly at the time unused by any known Apple product?” he asked. 
Well, one reason might be that Apple sold 70 million ARM devices in the last year and a half. The iPod, as everyone seems to forget, uses the ARM processor, or in most cases, two of them. Apple develops software for the iPod.
It appears clear that the iPhone itself also uses an ARM processor. For this reason, Gardiner says, “LLVM may allow developers to use Java and other languages, which are allegedly locked-out on the iPhone.”
Except that Java is not “allegedly locked out” of the iPhone as part of some accidental oversight that LLVM would help to solve. A Java VM isn’t allegedly missing, it is purposely missing.
There is no Java VM on the iPhone because it would unnecessarily duplicate Apple's own Cocoa environment with a large footprint, generic development environment that would double Apple's efforts while providing negligible benefit to anyone apart from Sun, who could at least crow that the iPhone has Java on it.
Again, the iPhone needs Java like the original Mac needed DOS, or like a unicycle needs an extra wheel or a set of skis. In addition to being a extraneous and superfluous addition to the iPhone, it’s also not something that would deliver apps of iPhone caliber, because Java is a lowest common denominator solution in the mobile space.
Which Brings Us Back to... D'oh!
Speculation aside, the future prospects for third party software on the iPhone all comes down to Apple deciding that third parties will contribute more to the iPhone than than they cause in troubles.
That makes it unlikely that homebrew development will come to the iPhone anytime soon, apart from the opportunities offered by widgets and within its Safari browser.
The hegemony of Windows on the PC desktop has similarly pushed many open developers--including Google--to a new platform: the web. Similarly, Sun built Java as a platform to enable software development that would result in code that could run on its server hardware. With no alternatives to Windows, Sun would have had few opportunities left to sell its servers.
The adaptations made by Sun, Google, and other developers to live in a world dominated by Windows offers some lessons to those who fear there will be no alternatives to Apple's $5 iTunes titles.
Where there's a will, there's a way. Once Apple has an installed base of 10 million phones, is it at all likely that the company would maintain the iPhone as a closed platform? I don't think so. I have some other predictions for WWDC though, which are up next:
Like reading RoughlyDrafted? Share articles with your friends, link from your blog, and subscribe to my podcast!
Did I miss any details?
Haloscan Q107
Bookmark on Del.icio.us | 
Discuss on Reddit |
Critically review on NewsTrust
Forward to Friends | 
Get RSS Feed |
Download RSS Widget
Check out the Daily Show Multi-Pass on iTunes.com
Next Articles: 
This Series
An iPhone SDK? Predictions for WWDC 2007!
Saturday, June 9, 2007
Ad
