http://www.ubnt.com/challenge/
They’re offering $200,000 in prizes for new Web UIs for OpenWrt to run on their new router. That’s really cool, but from my point of view the X-Wrt stuff is actually pretty damn good. The thing holding back making open router platforms one of the cooler things around is nothing to do with the Web UI.
The real problem is two-fold, first the most supported hardware is ancient and barely capable of running anything useful and can’t make use of any the new, cool rapid development languages because they’re interpreted and their runtime is bigger than usefully fits in the memory on these devices.
Second—and related—the development for these devices is a royal pain in the ass. Because of the devices limitations, you are forced to use a compiled language and you have to actually compile it on your own machine which isn’t so bad, but it makes debugging and profiling really hard because you have to make the code, deploy it to the device, and then use whatever limited tools you can get to work there to figure out if things are working right. If they aren’t you have to repeat and it’s brutal.
My little time spent working with that stuff gave me a huge newfound respect for embedded device developers. What we really need is a better toolchain to build programs for devices like this. Also, if people really think that devices like this are going to play a bigger role going forward, maybe we need to spend a few extra bucks and get them some real storage and make them a bit beefier.
