[bodeadly]

This post has long since scrolled off the front page but for posterity here is the best part:

"The fact is that good system design takes time; it is the sort of thing that requires hard solo thinking along with long discussions with other engineers. There are days when no real progress seems to be made, and other days when the only progress is to realize that what you thought was progress over the previous few days or weeks was in fact a wrong turn that won’t really work. Such a realization is progress. In fact, such a realization may be the most important kind of progress, as it can save huge problems later in the project. But to a manager it may not seem to be moving forward."

I recently came to this exact realization. So when I saw this post I had to see if it was in there. And it is!