Mike Rowehl: This is Mobility

Of course, I’m going to need to weigh in on the Android/Open Handset Alliance news from yesterday, but I would like to take the whole discussion in a different direction. Tons of people have been hyping it already despite the fact that there’s no real info, and tons of people bashing it because there’s no real info. Although I’m tempted to say that just like their ability to get people to blog about a 404 last week, Google has just manged to pull another fast one. I’m not going to however. I’m totally onboard with the idea that open platforms are the way forward in mobile, something I mentioned again and again at Mobile 2.0. So I just can’t pass up an opportunity to forward the discussion some while attention is turned that way.

First off lets try to line up some perspective. Lets say you founded Danger, lived through a lot of the challenges there and learned from them, and then had your next startup snatched up by Google and found yourself at the command of vast resources and power, what would you do? The Sidekick line of devices gained a fanatically devoted following of users. Sidekick fans were like the iPhone fans of the day, their device could do no wrong. However the devices never really cross over and made it mainstream enough that Danger enjoys a significant market share. Why? And what would you change if you were trying a second time?

The first and most major is the “closed platform” part. Most folks have a ton of respect for the massive distributed platform that was the Danger backend. Their ability to deploy new applications or revisions to applications from a central location and have it show up on all their devices, that’s just plain hot. However the setup didn’t encourage or enable third party developers really. Apparently it was somewhat possible to develop for the Sidekick, but I never got a good understanding of what was going on there. At the minimum it seemed to require you talk to Danger or T-Mo if you wanted to deploy your app. If it was me I would try for the same kind of “magic services in the cloud” idea that Danger had, but in a way that allowed anyone to deploy their application against the platform. One way of doing that would be to provide a generic runtime on the device that downloaded a client plugin that could run in a virtual machine, and services to hook into the base platform on the device and communicate with the network. The services facing the device would need to be stuff like alerting the user and providing different views of data so that any app can expose “homescreen” functions. The network facing functions would be stuff like updating the service transparently when a new version is available and background data transfer and syncing. Is that what’s behind this SDK that we’re going to see next week? I have no idea, but if I were Andy that’s the direction in which I would be thinking.

The second major takeaway that I would have from Danger is that device distribution can kill your market even if you have a fantastic service and a great device. One way to solve that is by forming a partnership with a number of manufacturers who already sell devices through a number of top tier carriers in all the major markets instead of going it on your own. In that respect HTC, Samsung, Motorola, and LG don’t sound like a horrible mix. Sure, none of them individually are dominating the handset market. None of them are willing to throw their support 100 percent behind the platform right off the bat. But so what? Together they represent a channel into just about any market, and that solves the distribution problem pretty well. I see no problem with this, at all. Apple isn’t there? Nokia isn’t there? Don’t care. Doesn’t make a difference.

The final part, this whole bit about being “open”. I’m an open source bigot, I admit it. But I’m trying not to lean too much one way or the other on this cause there just isn’t enough info. So what would have to happen here for this effort to not just be another hat tossed into the ring of open mobile (like OpenMoko, Ubuntu Mobile, LiMo, Qtopia, etc). It would have to be possible for other folks to implement the OHA standards and participate in the platform without having to use the stack top to bottom. If I could pick up the SDK we get presented with next week and start hacking together some scripts for my N800 to experiment with the services, perfect. It has to be possible for the OpenMoko folks to go “Cool, we want to throw our support behind this too!” and just go off and implement something that makes them OHA compliant and recognized as such. If implementing against a shared public unprotected standard makes you part of the Open Handset Alliance, perfect, we’re seeing some good. If participating in the alliance requires anything else except consistently contributing good code, it’s not really sticking to open principles and probably won’t really make an impact.

This entry was posted on Tuesday, November 6th, 2007 at 2:31 pm and is filed under Community, Open Source, ThisIsMobility. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.