iPhone 3.0 is on the horizon; Symbian has stolen a march on rivals with its new ‘application warehouse’ approach; even Windows Mobile is gaining new features. But Google’s Android continues to grab the headlines, as the scale of the search giant’s mobile software ambitions become clearer.
Real Android products may be thin on the ground, belying Google CEO Eric Schmidt’s promise of a ‘very, very good year’ for the platform, but it is spreading its tentacles into a host of non-handset devices too, foreshadowing the day when the cellphone and notebook are just two among many gadgets connected wirelessly to the internet. And an apparently routine update to mobile Gmail also points to even more disruptive software strategies that could push Google’s vision of the mobile web model to the forefront.
In the short term, though, Android is having little real impact on the market. Schmidt pinned high hopes on Android as Google reported slowing revenues and sought to calm investor jitters. The company reported first quarter net income of $1.42bn, an 8% year-on-year increase, and beating Wall Street forecasts partly because of a recent cost cutting program. Revenue rose just 6% to $5.51bn, and was also down 3% on the fourth quarter of 2008, Google’s first quarter-on-quarter decline since it went public five years ago. The main culprit was the cut in online advertising spend. But while the CEO emphasized Android as a growth opportunity, he failed to offer specific sales figures or targets for the firm’s mobile activities. So far, it remains unclear where exactly all this success is going to come from, with Sony Ericsson and Motorola both cautious on timescales for their Android launches.
As if the market weren’t frustrated enough with the uncertainty about the shipment of another Linux-based smartphone, Palm Pre, Sony Ericsson is now trying to manage expectations about the release of its first Android models, which should be a much needed boost to the venture’s flagging fortunes. CEO Hideki Komiyama told news agency Reuters: “It does require a lot of evaluation, as well as a lot of testing, a lot of acceptance from a consumer viewpoint, and there is still some time to go’. Third quarter still seems most likely, though, if comments made by Orange are to be believed ‘ the firm expressed hopes it would launch a Sony Ericsson Android device in Q3, raising speculation that the French giant might get an exclusive in key territories like France and the UK.
Komiyama’s comments highlight two important challenges for Android that may slow down initial uptake. One, the sheer difficulty of getting a brand new handset platform market-ready, on multiple devices, to the satisfaction of a huge number of network operators. Google has always had hugely ambitious timelines for this, even though it is competing with systems like Symbian that have been developing for years, and has often been accused of underestimating the complexity of getting each product certified by operators ‘ one issue that was reported to have delayed a launched planned for 2008 at China Mobile.
Two, there is the inevitable downside of open standards, the problems for vendors in creating competitive edge. “Our focus is on how can we differentiate from the competitors using the same operating system,” Komiyama said. His firm will focus on the user experience and interface, and its Xperia X1 Windows phone has already demonstrated what it can do in this respect ‘ but its rivals will be taking a similar approach, and cutting edge will be hard and expensive to achieve.
Nevertheless, HTC’s second Android phone, Magic, is about to ship at Vodafone and T-Mobile USA, and Samsung plans a launch in mid-year. “There are announcements happening between now and the end of the year that are quite significant from operators and new hardware partners in the Android space, which I won’t pre-announce except to say that they really do fulfil much of the vision that we laid out more than a year ago,” Schmidt said. But of course, market leader Nokia will not support Android and LG has put its primary focus on Windows, making launches from the rest of the top five ‘ Samsung, Sony Ericsson and Motorola (also hazy on timescales for its make-or-break Android plan) critical to gaining critical mass in the near future.
The complexity of the handset development process will explain the accelerated efforts to put Android on other types of devices, including set-top boxes and netbooks. In particular, the Japanese consumer electronics sector is showing rising interest. A group of Japanese CE manufacturers, calling themselves the Open Embedded Software Foundation (OESF), is looking to use Android in a range of devices with embedded wireless, such as set-tops, VoIP phones, karaoke machines, security and monitoring systems, and digital photo frames. The founder members are ARM, KDDI, Japan Cable Laboratories, Alpine Electronics and Fujitsu Software, and first products should appear at Japan’s largest electronics show, CEATEC, in the fall. JVC is also said to be developing an Android TV set.
OESF chairman Masataka Miura said the Foundation had been established in February to create a viable Android platform for embedded products and now has 25 members. It will look into applications and services, as well as marketing and awareness programs, not just hardware. While ARM has stolen a march on Intel in being an original member, Miura told EETimes that many semiconductor companies are also interested in joining, including Intel itself, TI, Marvell, Freescale, Qualcomm and Renesas. The group plans to open offices in Taiwan and South Korea this summer.
An important aspect of Android’s push into all kinds of low power devices will be its increasingly close partnership with ARM, the main supplier of processor designs for mobile gadgets, and ARM in turn has an important and close relationship with Adobe, which wants to position its Flash technology as a unifying force across PCs, cellphones and other devices, and across multiple operating systems. Adobe is working on Android implementations of Flash and other key technologies, and also on low power variants based on ARM. And now it has announced a new version of Flash that puts the system into internet connected TVs, set-top boxes, Blu-Ray players and other digital home products. The CE optimized version of Flash will allow viewers to see high definition video, interactive applications and new user interfaces on their TVs. Although Intel is an initial supporter, look for ARM to sign up soon ‘ one of the key problems vendors are finding as they start to develop low power ARM/Linux netbooks is poor support for high end video on the processor, as well as a fragmented software base.
Meanwhile, apparently minor enhancements to the Gmail messaging system may point to wider Google moves, that could have an even more disruptive effect on the mobile web experience and business model.
The type of underlying technology powering a new user interface upgrade for mobile Gmail could sound a deathknell for specialized downloadable clients, optimized for specific devices or networks ‘ such as the Skype clients for 3/INQ, Nokia N Series, BlackBerry and iPhone. These offer greater efficiency and ease of use than most pure browser approaches on current devices, but threaten a new kind of fragmentation and lock-in. Google’s vision of the ‘everywhere internet’ relies on running apps from any browser on any device or network, and it has taken some steps in this direction. Its new technology will make it easier to develop robust mobile web applications without having to create individual clients, said the search giant.
While the short term aim is to speed up email actions and support poor network connections ‘ using HTML5 for low quality connections, or Google Gears when offline entirely ‘ the MIT Technology Review points out that the Gmail upgrade is one of the first successful attempts to create browser-based applications that look and feel like those downloaded directly to devices, which could turn the current assumptions about sales and distribution of mobile software on their head.
Against this backdrop, a response from Nokia is widely expected. Of course, the main one has come from the open sourcing of Symbian and Series 60 to provide a more established and mature open source option for handsets. However, Nokia does use the Maemo Linux platform in a few products too, notably its Internet Tablets, and with its recent heavy hints about a more aggressive cross-platform strategy, is likely to use this system to drive a push into netbooks and other non-handset devices. This could entail the porting of Series 60 to Linux, and will certainly see Nokia pouring R&D resources into beefing up its implementation of Maemo to act as a viable counterweight to Android in the Linux world. All of which could make Mobile Linux even more fragmented than it already is.
Of course, Symbian will remain the key string to Nokia’s bow and the Symbian Foundation, which is responsible for the open sourcing of the platform, has detailed its concept of an ‘application warehouse’, which could provide a measure of unity across the proliferating app stores and mobile operating systems.
The Foundation said earlier this month that it would be taking a new approach to software distribution, contrasting with that of vendor or operator stores (including Nokia’s own Ovi Store of course) that are tied to one product or network. Instead, Symbian will supply multiple stores with applications written for its OS, which has just started the long process of being fully opened up (the deadline its June 2010, but so far only 4% of the code packages have been opened). Much of the work will be done not by the Foundation’s 200-strong workforce but by the open source community ‘ though, as many observers point out, ‘the community’ may equate, especially in the early days, to Nokia’s software teams.
For the apps warehouse, the Foundation will offer standard APIs, which third party stores can use to select and run the subset of products they wish to offer to their customers. The
Foundation will ‘sign’ the apps (nearly all mobile applications have to be signed in the interests of security, but there is constant wrangling over who should have signing rights ‘ operators or device makers). In Symbian’s plan, it will actually license third parties to sign apps and will create a repository to hold the apps. This would only be available to virtual shopfront partners, not the general public, and will not make any money ‘ it will purely exist to encourage Symbian developers and partners, by aggregating a large base of apps, and so making it attractive to operators. The Foundation will just take enough money to cover its costs.
As well as Symbian in open source land, there is LiMO and also Intel, which will push its Moblin Linux OS into a wider range of devices as it extends the reach of its Atom low power processor beyond netbooks and into smartphones, embedded devices and so on. With next year’s iteration of Atom, codenamed Moorestown, will come a new version of Moblin, which Intel recently put into an open Linux foundation, indicating its desire to get broader uptake of the platform, potentially at the expense of Android.
Moorestown will be an energy efficient chipset aimed at mobile internet devices, in competition with Qualcomm Snapdragon and other ARM-based designs. The chip family will be based on the company’s Lincroft system-on-chip, which includes an Atom microprocessor core and a memory controller hub, and will support WiMAX and/or HSPA. The accompanying chipset is codenamed Langwell. The new version of Moblin software will be optimized for a PC-like experience on the mobile internet and cellular voice capabilities, Intel said in a statement.
So Android will not have everything its own way, and there are still many warning bells to sound. Part of its appeal, apart from being open and free, is that it is ‘really just an embedded kernel’, as analysts at In-Stat put it, with the rest of the software stack, from user interface to browser to apps, playing a more important role than in traditional handset platforms. This allows vendors to target Android fairly easily to a wide range of different devices and usage models, making the difference in software; but it also comes with the downside of potential fragmentation.
Vendors have to do a lot of work in software to create fully robust browsers and applications ‘ clearly Google hopes they will take a short cut and use its own offerings, which have a clear headstart in terms of integration with the OS. But if they do not, they will have to work hard to optimize applications for the system, which lacks many native capabilities as yet. “Android today is still very basic and rudimentary. As an early stage product, it still needs to improve to be a truly commercially usable open source platform,” Chris Fisher, president of zoomMediaPlus, told EETimes. Until then, CE suppliers need to “take Android, customize it and make it their own’.
“Google and the Android folks have a real challenge in the market on their hands,” agreed Jack Gold, an analyst at J Gold Associates. “They basically have to become more than equivalent to what’s already out there.”