Re: About the proect of prioritizing a task list

[Robert <http://profiles.yahoo.com/robertb>]

 > From: Bill Bruns <http://profiles.yahoo.com/billbruns>
 > Date: Mon, 17 Jul 2017 00:28:16 +0000 (UTC)
 >
 > Hello Robert,
 > about the project you mentioned yesterday,There are at least two important 
 > dimensions to consider in determining the priority of a task:Importance, and 
 > Urgency.
 > Seems to me it would take a system a "long" time to learn these two 
 > characteristics if it is only given the task descriptions, with no information 
 > about due dates, or other clues about urgency versus importance.
 > The project seems to need more definition.
 > Do you have more to say on this?
 > Â Bill Bruns

Agreed.

I think that certain recurring tasks should be given appropriate priority.
Perhaps those could fall outside of the "importance" of other tasks.

Some meta data, such as due dates and recurrence frequency, could be
provided to the model.  But, I, as a user, shouldn't need to know those
per se.  All that I should need to know if that the top of my todo list
needs to be done before the stuff that follows.  (It's possible that the
system would just "learn" that these are important because they appear to
get done before other tasks when they do come up.  It's probably worth
trying without the meta data and see if it figures it out on its own.)

For example, I have this on my todo list:

 set "sleep" on primary MP3 player
 pay $91.66 June 2017 Credo bill
 change water in bike water bottle
 download podcasts to secondary MP3 device

Because I end up doing 'pay $91.66 June 2017 Credo bill' before 
'set "sleep" on primary MP3 player', the next time I get a similar task,
e.g., 'pay $91.66 July 2017 Credo bill', it should know that that task
should come before 'set "sleep" on primary MP3 player'.  In this way, the
tasks which get completed most promptly end up at or near the top of my
todo list.  (This implies both an "importance" and "urgency" of the task.)