Counting seconds like that isn't exactly correct in my experience. I find that I actually don't need to go that fast on the whole. For instance, reading HTML d

Counting seconds like that isn't exactly correct in my experience. I find that I actually don't need to go that fast on the whole. For instance, reading HTML docs in a browser with a text editor vs using a fancy IDE. Metrics say IDE wins, but I don't think those seconds shaved are in the critical path. And there are other benefits to the non-IDE workflow.