Mark Forster proposes a will-do list instead of a to-do list. Pick one day's worth of items that you'll work on today and do them. If new stuff shows up, schedule them for tomorrow (or later). The will-do list is a closed list: nothing will be added. That way it is actually possible to complete it. You'll never manage to kill off your entire 200-item to-do list, right? (Incoming emergencies are marked especially as "this came in during the day").
In omnifocus (my macintosh GTD app of choice) I'm using the option to flag items for this. Every morning (or on the previous evening) I go through the avaiable tasks and flag a few of them that I'd like to accomplish today.
Problem: I flag too many items, so they currently drown out the rest. On the one hand: fine that I actually do what I'm planning to do. On the other hand: I depend largly on a quick scan of available tasks early in the morning. There's a risk of working too much on seemingly important items in a wide variety of projects: lack of focus and possible lack of working on the really important things.
I'll have to try and cut down on the amount of flagged items: use them more selectively. That way they'll also keep their power. Or I ought to be more focused on certain projects when choosing the items. Anyway: not perfect, but it sure does help me.
I borrowed his book (Do It Tomorrow and Other Secrets of Time Management
My name is Reinout van Rees and I work a lot with Python (programming language) and Django (website framework). I live in The Netherlands and I'm happily married to Annie van Rees-Kooiman.
Most of my website content is in my weblog. You can keep up to date by subscribing to the automatic feeds (for instance with Google reader):
