SavaJe wins LG mobile OS gig
While Palmsource points to Linux
SavaJe, the start-up mobile OS maker heavily backed by the cellcos, has won its first major handset supporter, LG, an alliance that fits well with the Korean manufacturer's strategy of keeping close to the large operators to win western markets. This would have been more serious for Nokia two years ago than it is now - the Finnish giant is increasingly looking to dominance of the user interface and the higher software layers, rather than the base OS, for its power, appealing to developers on the same level as Microsoft. The same thought has occurred to PalmSource, which aims to preserve its life by bringing the advantages of its software environment to the Linux platform.
SavaJe, the Java-based mobile operating system heavily backed by major cellcos, has gained its first partner from the top five handset makers, LG Electronics. Like most developments concerning SavaJe, politics will have played a greater role in the decision than technological considerations. LG's move emphasizes its bid to acquire major market share in the west by positioning itself as the operators' friend, and taking advantage, as a relative newcomer, of the deep seated tensions between the large carriers and the handset giants, particularly Nokia.
Nokia itself believes the war will be won on the strengths of the user interface environment rather than the raw OS, and so is focusing increasingly strongly on establishing Series 60 as a de facto standard - an area where the cellcos lack tame technology. PalmSource sees the wisdom of this approach, and this week made plans to fill the gaping hole in mobile Linux, a strong user interface, with its own technology.
The SavaJe operating system has significant support from the major European cellcos, which favor it for three reasons - since it comes from a start-up, which they help to fund, they can control it; it offers a highly customizable user interface, vital to the operators' branding and content delivery strategies; and it is a useful stick with which to beat Nokia and the other handset makers.
All these factors were highlighted by the formation of the Open Mobile Terminal Platform Alliance (OMTA) in June by Vodafone, NTT DoCoMo and most major European cellcos. The group will specify a design framework and programming interfaces for cellphones and will issue a compliance test. The implication is that non-compliant devices will be excluded from the buying plans of the operators responsible for a huge proportion of phone purchasing outside the Americas - putting massive pressure on the handset makers to comply. This effectively shifts the balance away from the traditional and much resented relationship between phonemaker and operator, in which the former dictated design and had the key branding.
Now the cellcos want to establish their own brands and to drive design to their own agenda - boosting average revenue per user by delivering rapidly changing content services through a highly distinctive user interface; in so doing, increasing customer loyalty and attracting more of the user's mobile spending to the operator-delivered content, and less to the handset itself, which the cellco has to subsidize.
SavaJe, despite some technological shortcomings, does support the operators' desire for more individual user environments on the phone through its native Java architecture. It is malleable, being dependent on the operators' support for survival, and being 26 per cent owned by Vodafone, Orange and T-Mobile. And it represents a scarcely veiled threat to Nokia and the others - fall into line, or SavaJe will become the preferred platform, ahead of Symbian OS, Windows Mobile and other vendor-driven solutions.
Against this backdrop, it is unsurprising that the first handset maker to break ranks and support SavaJe is LG. It has moved into the top five global handset makers, and the number one spot in 3G phones, by playing the cellcos' game - something it is well accustomed to doing in its native Korea, where the operators have the upper hand in driving and specifying handset features. This has made LG, like some second tier Asian manufacturers such as Sharp, highly attractive to the western operators. They have greater creativity and R&D resources than the pure white label phonemakers, but still give the carriers significant input into the design process.
SavaJe itself remains unproven, but both Texas Instruments and Intel have optimized their cellphone platforms for it, in anticipation that the backing of the major carriers would soon persuade some major handset companies to give in and include it in their portfolios. It badly needs to be tried out in the real world, and should appear in several LG GPRS and 3G handsets in the early part of next year.
So far, LG's operating system-neutral stance (it will also deploy Symbian/Series 60 phones next year) and carrier friendliness has given it an early headstart in the fledgling European 3G market. As Vodafone and Nokia bickered over what the 3G handset should look like, LG was creating usable designs. Although only three per cent of mobile phones sold in the EMEA region in the third quarter were 3G compatible, according to Canalys, LG took a 55 per cent market share.
The upper layers
While operator backing will ensure some market share for SavaJe, it represents less of a threat to Nokia than it would have done a few years ago. This is not just because it is still immature and has technical shortcomings compared to Symbian - such as a large footprint - but because the sphere of influence is moving from the OS to the upper layers. Java itself is now far more important than the individual OS in differentiating handsets, and attempts to set de facto standards are revolving around user interfaces and development platforms - Nokia's Series 60 being a prime example. If Nokia can make Series 60 sufficiently dominant among phonemakers and developers, it will be hard for operators to ignore - and at that stage, the Finnish giant might consider porting it to other OSs, even SavaJe itself.
That would have been inconceivable a few years ago, when the battle to dominate the emerging smartphone platform was firmly focused on the capabilities of the OS. The point is emphasized by another software house defocusing on its operating system, in this case PalmSource, which is to create a Linux version of its software platform.
PalmSource goes Linux
In the short term, this move - unthinkable when the software house was still part of Palm - is designed to gain market share in China, the most enthusiastic country about mobile Linux. To this end, PalmSource is to acquire developer China MobileSoft, which has been developing a mobile version of Linux for smartphones as an alternative to the dominant variant in this market, MontaVista Linux.
This is not just about China however. The PalmOS, though still strong in market share on enterprise focused PDAs, clearly has limited shelf life in the smartphone arena. Its largest customer, PalmOne, recently indicated that it would develop devices for other operating systems and there are few other licensees, following Sony's decision to withdraw the PalmOS-based Clie handheld outside Japan.
PalmSource, then, needs to defocus on the OS and apply its technological strengths to more mainstream platforms. Palm achieved its initial success in the PDA sector by building a strong developer community with attractive tools and user interfaces, and these advantages can be applied to Linux, which lacks such benefits in the mobile world.
The move is a clever one that mirrors Nokia's efforts to gain influence on the broader community through Series 60. The main problem for Linux on phones is its lack of a strong user interface environment. This means that creating an attractive handset for users and operators requires far more development effort than is typical for the higher level OSs such as Symbian OS with Series 60, or Windows Mobile. By deploying the Palm OS upper layers on top of Linux, the problem is solved and PalmSource has a whole new base in which to sell its software and potentially create a de facto standard.
"It may not be clear to everyone just how great a potential is being unveiled today," said Bruce Perens, co-founder of the Open Source Initiative. "Much of our existing desktop and server software can be moved to the mobile platform just by typing 'make' and waiting for the sources to compile. That brings a power to the platform that lacks only a good user interface, PalmSource's forte."
Drawing on China MobileSoft's technology, PalmSource will create a hybrid that has the look and feel and developer strengths of PalmOS, but on a Linux kernel. CEO David Nagel said: "We believe the combination of PalmSource, CMS and Linux gives us the technological and market critical mass to compete with even the biggest proprietary operating system companies."
For the time being, PalmSource will continue to offer its own two operating systems, PalmOS version 5.x or Garnet, and version 6.x or Cobalt.
Copyright © 2004, Wireless Watch
Wireless Watch is published by Rethink Research, a London-based IT publishing and consulting firm. This weekly newsletter delivers in-depth analysis and market research of mobile and wireless for business. Subscription details are here.