Why Apple can’t be too worried about Android 3.0 Honeycomb tablets taking away iPad sales: Part 2
February 8th, 2011
The world’s biggest software platform
Daniel Eran Dilger
Listen to giddy Android enthusiasts and you might get the impression that the next tablet-centric version of Google’s Android platform, named 3.0 Honeycomb, is about to destroy iPad sales. They’re wrong, here’s why, part two.
Part 2: The world’s biggest software platform
As noted in part 1, Android’s inability in 2010 to in any way damper sales of Apple’s iPhone (and even mount a credible alternative to the iPod touch or iPad) resulted in the platform simply supplanting what used to be the world’s largest smartphone platform. No, not Symbian. The world’s largest smartphone platform has long been Sun’s JavaME (“micro edition”, formerly known as J2ME), now owned by Oracle.
Prior to the Year of Android, JavaME served the common software platform for most smartphones, from RIM’s BlackBerry (which hosts the Java Virtual Machine on RIM’s proprietary kernel) to LG and Samsung models (that similarly hosted a JVM atop proprietary kernels in embedded products such as the LG enV) to Danger (which spawned the Android project later acquired by Google, but before that produced a line of Java-based Hiptop/Sidekick devices acquired by Microsoft) to Nokia’s Symbian (once also used by Sony Ericsson and Japan’s DoCoMo), Palm OS, and Microsoft’s Windows Mobile (all three of which host a JVM as an alternative to their own parallel native development platforms, similar to the Java support on Windows or Mac OS X).
Nobody ever talked about how JavaME “rivaled” Windows Mobile or Palm OS or Symbian; instead, firms like Gartner told us that Windows Mobile was poised to make headway in pushing its own native APIs at the expense of Java, while RIM boasted in its own peripheral services which enhanced its version of the Java platform among enterprise users, and Danger promoted its Java-based phones as being popular among text-crazed youths, and so on.
Why Apple can’t be too worried about Android 3.0 Honeycomb tablets taking away iPad sales: Part 1
Why Apple can’t be too worried about Android 3.0 Honeycomb tablets taking away iPad sales: Part 2
Why Apple can’t be too worried about Android 3.0 Honeycomb tablets taking away iPad sales: Part 3
Apple takes on the monoculture status quo
It wasn’t until the debut of iPhone that a significant new smartphone vendor was willing to point out that JavaME’s empire was embarrassingly naked. Despite purporting to be a “write once, run anywhere” platform, JavaME in practice was a “develop software, then adjust it to work on every specific phone you want to be able to deploy it on,” a plague reminiscent of Java on the desktop but far worse because of the additional complications inherent in mobile devices.
While there were plenty of JavaME apps, getting them to work well on any specific phone was simply orders of magnitude harder than building complex web apps that could run flawlessly across all the different web browsers without resorting to Adobe Flash in the days before HTML5.
Apple’s solution was to simply not include Java, software which in reality just replicated the features of its own Cocoa APIs without potentially enabling the iPhone to run any significant, useful library of quality mobile software. Apple’s purposeful omission of Java was such a non-even that nobody made much mention of it, because nobody thought that running Java mobile apps was really much of a feature. The lack of Java support on iPhone was completely overshadowed by its similar lack of support for Flash, because users actually saw some reason why one might want to access Flash content.
Additionally, the lack of Java support on the iPhone made Cocoa the only game in town for building apps, creating a funnel of development that resulted in a library of superior, new apps that were consistent and attractive, in stark contrast to the often ugly Java mobile applet junkware that existed in buckets on the web already. This worked along the lines of the 1998 iMac dumping old legacy ports for USB, resulting in an instant market for USB peripherals.
Further, just as PC makers prior to the iMac failed to see the mess of legacy ports as a real problem and were even slow to decisively embrace USB afterward, mobile makers failed to recognize that JavaME wasn’t really very valuable. Dial back the web to 2006, and you can read all sorts of reports about how everyone was declaring their support for Java and promoting it as a bullet point feature, as if smartphone users were really benefitting from the potential to run ugly junkware, poorly.
The iPhone Threat to Adobe, Microsoft, Sun, Real, BREW, Symbian (2007)
Sun Tries to Jump on iPhone Bandwagon with jPhone (2007)
iPhone 2.0 SDK: Java on the iPhone? (2008)
So that’s why Oracle is pissed at Google
Google took a different approach than Apple, because it didn’t have its own mature development APIs to use in building a mobile operating system as Apple did. It had started with Android, a licensed Java platform originating with the developers of Danger. Rather than building its own new platform as Apple did, or as Microsoft had with Windows Mobile, or as Palm would later do with webOS, Google simply modified the Java platform (starting with Apache Harmony as its class library and pruning off compatibility with JavaME) until it got to the point where it decided it wouldn’t have to license any underlying technology from Sun in order to distribute it for free.
Sun appeared to look the other way as this happened, but after Oracle acquired Sun, it accused Google of stomping on Sun’s intellectual property to deliver Android, and has since filed suit with the intent of collecting royalties from Google’s modified version of the Java platform. While that legal argument is being heard by the court, there is a clear and obvious result of Google’s actions that is uncontroversial: Android has virtually replaced at least half the market for JavaME in a very short amount of time, and threatens to squash a large portion of what’s left.
This is bad news for Oracle/Sun and BlackBerry (and Danger, but after Microsoft’s acquisition and the KIN fiasco, the group doesn’t really have prospects for good news) because it means the world’s supply of developers interested in JavaME is bound to dry up and blow away. It’s sort of good news for LG and Samsung and Motorola and HTC and Sony Ericsson, because it means that firstly, they don’t have to pay for JavaME licensing anymore (at least until Oracle wins a claim against Android), and secondly, they they don’t have to manage their own JVM implementations, because Google supplies them with free code that is more likely to work without nearly as many issues as the scores of slightly different JVMs that once existed.
Android’s replacement of JavaME is therefore a lot like Internet Explorer’s replacement of Netscape as the world’s web browser around 1997: a bigger, more powerful vendor offering free software that replaces something that had a lot more flaws and inconsistencies under the hood, with promises that it could deliver faster, better progress that would better facilitate interoperability. When the entire world shifted to IE, it did make it easier to deliver web sites that worked more reliably. With Android, it is similarly quite a bit easier to roll out apps than it was in the JavaME universe.
Google has therefore all but destroyed the JavaME platform and replaced it with its superior alternative. However, this has had nearly no impact upon Apple, because unlike RIM, it does not benefit from Java development, and unlike Oracle, it did not make money from Java. Apple’s iOS was already facing weak competition from Java in apps; now, that competition has been replaced by Android licensees. Despite this change, Apple’s development platform is not at risk because like JavaME, Android offers very little in terms of commercial incentives to actually build apps for its platform.
We have always been at war with Open-sia
Once again, just to emphasize that point: Apple’s iPhone took on the monoculture of JavaME and won (in terms of popularity, profits, app platform success, and every other metric). Nobody cared. Google has now resurrected the corpse of JavaME in a new incarnation that is threatening to reclaim a similarly large share among all non-Apple handsets (albeit without actually profiting from this, and without actually delivering a successful app platform that benefits users or developers), and it is hailed as the second coming of Christ and the needed comeuppance to Apple’s evil success story.
Cheering for Android is a bit like cheering for a desktop version of Linux struggling to replicate the brain dead look of the Windows desktop to take on Mac OS X in a battle for market share, when a) Apple has never gunned solely for market share, but rather for a profitable market segment that will enable it to continue to deliver the kind of products it wants and b) the “openness” angle at the core of this cheering appears to forget that Apple itself is leveraging open software within its core platform, while Google’s value layers are not.
The highest level open software in Android is Apple’s WebKit browser (also shared with Nokia, BlackBerry and Palm), while Android’s “with Google” apps are closed source commercial software. Android’s highly regarded Gmail app is not only closed source, but doesn’t even support interoperability with open mail standards, forcing users to resort to the second rate, separate but not equal “other mail” app to access email from competitors. That’s the kind of behavior one would expect from a Microsoft, not a Google, and the sort of thing that would be reviled as “EmailGate” were it occurring on the iPhone.
Google similarly seeks to disrupt open interoperability in promoting WebM as a “royalty free” alternative to H.264, an effort that really just pushes Adobe Flash at the expense of HTML5 video. In parallel, it also supports SMS-based messaging within Android for push notifications and silent software installation (SMS is a fee based telecom standard), while Apple built its iOS push notifications upon open standards (XMPP/Jabber PubSub) and similarly built AirPrint as a feature of CUPS, the free printing architecture owned by Apple and freely shared with Linux and other Unix distributions).
Who again is the hero and the villain of open interoperability? There seems to be some mismatch between words and deeds. According to Android advocates, the difference between Apple and Google is a simple matter of black vs white, but this sort of ignorant, fact-free campaigning on buzzwords that do not reflect one’s actual actions is only serving to give Google carte blanche to pull the wool from very black sheep over the eyes of those who choose not to look at reality.
A staged battle of dramatic conquest
Curiously however, Android isn’t being portrayed in the media as rivaling JavaME (which it certainly has), but as threatening Apple (which it really hasn’t). This isn’t so much of a mean spirited conspiracy against Apple as it is just an attempt to spin a dramatic story of conflict.
Apple and Android have only ever served as rivals in a symbolic sense. Google had always been gunning at taking out Windows Mobile and herding the JavaME (and Linux) cats into line in a way that Google could benefit from as an ad vendor. Apple, other the other hand, had been aiming at producing a desirable smartphone people would want to buy. These two objectives only ran into each other in minor, peripheral aspects.
Of course, Google has appropriated Apple’s game plan, shifting Android from being a BlackBerry/Windows Mobile killer to being an iPhone alternative, but as overall strategies, the two have very different objectives, as is obvious from Google’s business model revolving around creating a new display surface for ads, and Apple’s business model that revolves around hardware sales, accessory licensing, app sales, media sales, iAd app monetization, retail sales, cloud subscriptions, and related desktop, notebook, tablet, and media player sales.
The fact is that both Apple’s iPhone and Google’s Android platform can grow tremendously while only ideologically battling in the dramatically staged war zone of blogs and fan-sites. Apple is certainly competing against HTC and Motorola and Samsung, but those skirmishes aren’t very interesting because Apple usually always wins in terms of units sold, user satisfaction, developer interest, and profitability. The only way to make the race watchable is to pit Apple against every other phone maker on earth, collectively, under the banner of Android.
This story is intellectually dishonest for a number of reasons, but primarily, why is “Google’s Android” credited with the unpaid use of its platform by third parties in addition to (occasionally) the non-compatible use of forked versions by rivals (such as China Mobile’s OPhone and the Tapas OS)? Nobody credits Apple with “owning” the “market” for WebKit browsers.
Instead, Google’s Chrome is pitted against Apple’s Safari when market share stories are printed. That’s as silly as calling Google’s Nexus S a rival to Samsung’s Galaxy S, when both are really the same thing with some minor tweaks. Should Google really be credited for reselling other people’s work and hailed as superior to the products its own branded versions are based on?
Did anyone start reporting what share of the PDF reader market Apple had “stolen” from Adobe Acrobat when it began bundling a free Preview app with Mac OS X?
Along those lines, we could be launching into a discussion of Chrome OS, if Google had been able to deliver that last fall as planned. Instead, lets’s discuss the near future of the Android-based tablets about to materialize in Part 3: The Honeycomb Tablet.