[emacs-wiki-discuss] On Planner feature growth (was Re: (Repost) Switching projects)

[Sacha Chua]

"Raymond Zeitler" <address@hidden> writes:

> I really see this growing into something that, when invoked, will
> prompt you for a planner and publishing directory pair from a list,
> with a default choice available.  At that point, it works the way

<grin> How do people remember to keep track of all these other
projects? Clearly there must be a main planner project that you can
use to M-x plan your day. It could then contain lisp:// links to
switch to particular planner projects (sorry, interactive functions
only as I haven't figured out a nice way of specifying arguments yet;
maybe we should add a button attribute to <lisp> tags). Then those
projects could have lisp:// tags to move back to your main project,
and keyboard shortcuts...

If you get all of this working, please document your use somewhere.
http://www.emacswiki.org/cgi-bin/wiki/PlannerModeMethods would be a
nice place. I find that descriptions of how people use Planner tend to
be very useful (and inspiring), and may even be more helpful than a
list of functions (which you can get by reading *.el anyway). In fact,
I daresay most of planner's value lies in the structures people build
around it rather than in its intrinsic code. In contrast, something
like w3m-el is pretty cool, but most people use it in a fairly
standard way. (Then again, I've changed my w3m keymap so extensively
that I'd be lost on the default config.)

> find-file works, with tab completion and a *Completions* buffer. It
> should also be easy to set up. I could attempt to write that, but
> I'm afraid people will laugh! :-) And if my boss found out that I'm
> wasting time writing Lisp code....

As long as you still get your work done... <laugh>

Working on Planner has really broadened my horizons. Several comments
on the recent Slashdot thread about the Getting Things Done
methodology question the usefulness of spending so much time
optimizing one's workflow, and it _is_ easy to spend more time
planning than actually doing anything.

One of the things I like about Planner is that it doesn't require that
much thought. It actually feels easier than maintaining a ~/TODO list,
or embedding /* TODO: ... */ comments in my code, or using a
commercial-level PIM. I can just hit a key, write a task or note, and
then go back to work.

Because we evolve new features as _we_ use them, and due to the nature
of personal information management we tend to use the features every
day, eventually we get so annoyed with clunky interfaces that we'll
smooth them down into almost automatic stuff. For example, I hadn't
thought of referring to dates with +n / -n until someone suggested it.
Now I use it almost all the time. Same goes for
planner-task-dates-favor-future-p and the fact that it's separate from
planner-expand-names-favor-future-p: it makes it just that much easier
to schedule tasks for the future without having to specify full dates
for dates in the past. I briefly thought about controlling
planner-expand-names-favor-future-p so that it wouldn't do that for
ambiguous dates that were too far off (and thus probably meant to be
in the past), but decided that that would be a cognitive load I'd
rather not have to deal with as a user. <grin>

I spend a fair bit of time hacking on Lisp code, but I feel that it's
more of enhancing Planner. It already does everything I need. Now I
just tweak it to do more of what I want! ^_^ Besides, working on
Planner is braintwistingly fun, and I love working with such a
wonderful community.

-- 
Sacha Chua <address@hidden> - Ateneo CS faculty geekette
interests: emacs, gnu/linux, making computer science education fun
http://sacha.free.net.ph/ - PGP Key ID: 0xE7FDF77C