What Larry needs to fix as CEO

I found this post by Xoogler Steve Lacy very insightful. He gives a great view of what a long time Google engineer finds frustrating about the development process within the company, from an NIH ("Not Invented Here") blindspot to the myth of 20% time. A lot of the processes I describe in IFL that worked well within a small startup have evidently morphed into cumbersome bureaucratic obstacles in today's much larger corporation. Well worth a read.