Paul Graham on how to make things in 6 steps
I like to find
1.simple solutions
2. to overlooked problems
3. that actively need to be solved
4. deliver them as informally as possible
5. starting with a very crude version 1 then
6. iterating rapidly