We had this group who had some embedded development boards.
We have MSP430 Launchpad, Beagleboard and could 8051 and arduino board from the net. We wanted to develop some application and in the process learn something. We are trained men in the embedded field, hence we believe that we could come up with some nice application.
We start scouring through the internet jungle. Forums and magazines are helpful, but they give construction and sometime leave nothing for imagination. All the angles of perception get covered in most cases with forums and discussions.
So we get stuck in the select mode: either do monkeying or wait for the eureka moment. We have a bunch of experts who are accessible, knowledgeable and could guide us anytime. But we wait for the idea moment!
We start by reading the manuals and start with LED blinking in MSP430. Yes! we get stuck. No light. And we ponder over the question of copy pasting some net given code or read the manual for the code.
Sample code is the foundation or building block of any project. We do the sample code. The light blinks. What further? Standard exercises of UART, LCD interface and other communication protocols or get (copy) ideas and do something incremental?
Finally stuck there. With monkeying vs innovating question.
Ajeya: The Invincible's Crucible 
Hi,
You consider some graphics application.
Try interfacing a Nokia 5110 LCD to the board.
Develop low level display drivers.
Then you could implement an open source graphics library. Research further and once you are confident and your basic graphics library is built, then you could think of incorporating various graphics optimization algos within the library and the driver. Ofcourse, as it is a CPU and not a GPU, that would not get you far enough. But it is fun to do it. I would love to involve myself also in your activities.
LikeLike