Google's Android mobile phone stack will end up fragmented into multiple versions, Symbian's research chief has claimed.
Speaking at the Handsets World conference in Berlin, David Wood claimed there will be "multiple Androids in two years' time". Symbian has previously dismissed mobile Linux as being too fragmented in its nature, but that was before Google's Open Handset Alliance (OHA) was formed or the LiMo Foundation really got going. Nonetheless, Wood reiterated his company's stance on Wednesday, claiming "mobile Linux fragments faster than it unifies".
"The reason the code will change is that people will have different requirements," he said. "People have been predicting the success of open source for some time [but] fragmentation is easy, and integration is hard." Wood went on to list the extra costs of mobile open source: the costs of debugging; the loss of intellectual property through "copyleft"; licensing costs for "numerous upper layers"; and "opportunity costs from not being the most innovative".
Google claimed in November last year that it had made members of the OHA sign a non-fragmentation agreement, in which they promised not to "modify [the Android code] in non-compatible ways". However, a spokesperson for the company told ZDNet.co.uk on Thursday that, as Android was to be open-sourced, "there is nothing to prevent fragmentation".
"The idea behind putting together the OHA was to create a critical mass of standard implementations of Android," said Google's spokesperson. "If that critical mass exists, the motivation to fragment decreases. The advantage is for developers to build cool applications for it, so creating an implementation that fragments to the point where applications don't work doesn't make sense. What is the motivation to fragment if the main draw for Android, from a consumer perspective, is that you get the applications and benefit from innovation?"
Other delegates at Handsets World were generally enthusiastic about Android's potential, in many cases saying they had higher hopes for Android than for the LiMo Foundation's nascent platform. While Android is a full mobile software stack, LiMo — a rival industry consortium to the OHA — is only developing a common middleware component, on top of which its members can then put proprietary applications.
Javier Gonzalez, the chief executive of Aquamobile, told ZDNet.co.uk that Android was "more attractive" than LiMo because it will allow his company to "get closer to the hardware features and get a good user experience". Aquamobile makes software to read digital watermarks on, for example, business cards that have been photographed through a phone's built-in camera.
"Part of the work we have done for Java is reusable, and the APIs from manufacturers will be more accessible," said Gonzalez, adding: "Android's architecture is designed to facilitate the development of new business models or [take those models] from the fixed to the mobile world — the big opportunity is to build services that can make money right away".
Fabrizio Capobianco, the chief executive of Funambol — which makes personal information managemen (PIM) and synchronisation software for handsets and has one of the largest mobile open-source developer communities — was one of many at the conference to peg Android's chances to Google's status. "Android stole the show for me, because Google is so big," he said. "It's a company with unlimited money." HTC Europe's vice president, Florian Seiche, also said Android had a better hope of avoiding fragmentation than its rivals because "there is a sense of someone driving it to happen".
Texas Instruments's director of marketing for wireless, Jean-Phillipe Fournier, was less certain about Google's project, saying he was "not sure Android would help" combat fragmentation. Aditya Kaul, a senior analyst in wireless technology for Pioneer Consulting, said however that a degree of fragmentation — certainly in terms of having rival factions such as the OHA and LiMo "promoting their own technologies" — "reduces the market but helps drive innovation".
