Software Everywhere — David Beers on Mobile Computing

Simon Brocklehurst made the inevitable comparison. How will JavaFX Mobile handsets compare to the iPhone?

In my view JavaFX Mobile is not so much trying to compete with the iPhone as it is trying to ride on its prodigious marketing coattails. First of all, the two products have totally different customers. Apple is selling a handset to end users this year. Sun is selling a software platform to device manufacturers and it probably won't show up on devices for 18-24 months. Sun's goal with the announcement is twofold: (1) attract the interest of second-tier ODMs that are thinking about how to develop iPhone knock-offs, and (2) start the process of reshaping its image from "enterprise server company" to "personal and consumer technology company." I say they are approaching second-tier ODMs because, like ACCESS, I see them targeting device manufacturers who don't have the resources to build and integrate their own Linux + Java software stacks like Motorola, Panasonic or NEC, and thus need something that's complete and largely pre-integrated.

Mobile Java developers will be all over these JavaFX phones... if Sun manages to sell the ODMs and carriers on the platform. That's a very big if. Currently, the only mobile phones that ship with the powerful CDC Java runtime that JavaFX Mobile needs are the high-end Nokia Communicators like the 9300 (no longer sold by any US carrier) and the new E90 (yet to be announced for a US carrier). Sun bought a lot of the technology in this platform from SavaJe, which got many rounds of funding over many years to create a super-Java phone platform like the one Sun just announced. SavaJe never found any buyers.

I think the reason for this is that the carriers are already having a hard time locking down lower-power MIDP handsets to their satisfaction. What do they want with a dramatically less constrained device unless it's going to sell at a high premium like a Palm or HTC smartphone? Sun implies that their platform would show up in feature phones, which is just a fantasy to make this announcement sound a lot more exciting than it really is for the developers attending JavaOne. I asked Sun's CTO whether the platform was developed with feedback from the carriers and also why he thought SavaJe failed. He walked all around these questions (especially the second one) indicating to me that this isn't something that the carriers are asking Sun to deliver any more than they wanted ODMs to bring them SavaJe phones.

The more I think about it, JavaFX Mobile does not make any business sense. When it comes to Linux + Java, Motorola is a lot more likely than Sun to get the mix right. Sorry Sun, I just don't think you can jump into this market just by slapping together a bunch of cool components that get developers jazzed: in the mobile handset market it's the wireless carriers not the developers that are in the driver's seat. If there is one thing I have learned as a mobile software developer it's that you ignore this at your peril.

Unless, of course, you are Steve Jobs. ;-)