Greetings AutoFocus testers, especially my fellow Macheads! And hello, Mark, and thanks for all the ideas you keep generating! Will AutoFocus eventually become the subject of a new book?
I've set up an electronic version of AF using TaskPaper, with "task" entries only and no "projects". I initially created several separate files as "pages" with one screen of entries each (no scrolling). Later I realized I could put everything in one file and tag one screen's worth of items with @Today, then use TP's tag filtering function to isolate those. This provides the option (which I haven't used, and which probably isn't kosher Autofocus) of starting the day by cherry-picking a screenful of tasks from throughout the list, rather than going through pages in chronological order. Maybe that would also be a way to feel a little more secure that tasks with deadlines aren't being neglected. On the other hand, an issue I've been wrestling with in recalibrating my own self-management systems is precisely how to keep tasks with deadlines from crowding out those that are inherently more important but don't have deadlines.
Autofocus seems more amenable than DIT to integration/hybridization with GTD -- tasks can be put on the AF list after they've been "processed"/"clarified" (the latter is the new terminology from _Making It All Work_), then dealt with according to the AF instructions. Alternatively, I can drop an item in AF even if it's not clarified yet, take action _by_ clarifying it whenever it "stands out" and then add the "clarified" action to the end of the list.
I have a feeling that AF's ability to naturally filter out for "dismissal" those tasks that never manage to "stand out" is going to be one of its biggest advantages, and something sorely missing from GTD. (Or, you could look at it as a way to create a GTD Someday/Maybe list at the back end rather than the front.)
This probably shouldn't be considered a flaw of Autofocus, but like Mel at
http://www.markforster.net/forum/post/622031 I find the system stalls when I take up tasks that I "feel like" working on for many hours, especially if I'm not finished by the end of the day. "Little and often" has never been my way of working -- I'm more of a head-down-and-bull-through-it kind of guy. I think this is an issue for me that goes beyond testing AF, and starting tomorrow I'll probably try setting a timer and allowing myself no more than (say) one hour to work on any given task before rescanning the list. The issue might be worth mentioning in the final published version of AutoFocus.
At least for the time being I've decided to stick with a separate list of first-thing-in-the-morning routine tasks -- mainly because experience has shown that, if I still have some of those things to finish and am not ready to go to the gym after dropping my daughter off at school, the probability that I will still make it to the gym at all that day drops drastically. The list consists of a dozen or so tasks that each take between a few seconds and five minutes or so to complete. These are things that are not quite on the level of "brush my teeth" but close. When I started trying Autofocus Monday I dropped those onto the list along with everything else, but they're things that I really do want/need to do every morning, and I was finding that rescanning the list after each action was slowing things down unnecessarily. Not wanting to ignore Mark's recommendations altogether, I've radically simplified the list and moved those items that really don't need to be done first thing in the morning into AutoFocus. So, my routine will be to do the morning routine when I get up, take my daughter to school and then head to the gym, then dive into Autofocus when I'm done there.