[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Orgmode] time profiling and other meta-analytic commands (was automatic

From: Samuel Wales
Subject: [Orgmode] time profiling and other meta-analytic commands (was automatic clocking)
Date: Sun, 23 Aug 2009 15:20:24 -0700

My idea, which might or might not meet the automatic clocking thread's
OP's need, should be called time profiling to avoid confusion.

The idea is to figure out where in the outline hierarchy you
spend your time, in an approximate fashion, much like
profiling code. This is done

  - without polluting the numerous entries with LOGBOOK drawers
  - without the need to clock in and out of hundreds of tasks
  - with exheritance (that is, times get summed up the hierarchy)

It is only one example of a general class of possible
commands that do meta-analysis of the way you use org.
Other members of this class might include

  1) rate measurement, where we display how rapidly you add
     tasks vs. dispose of them on average over a period of
     time.  For example, if you add 10 tasks per day, but
     only mark 5 tasks per day as done, your rate is 0.5.
     This could even be charted.  X axis is time.
  2) design element utilization, where we display,
     numerically, how often you use each todo keyword, tag,
     priority, etc.  If you almost never use a todo keyword,
     it might or might not be useful.  You might discover
     that you are being lazy by marking everything TODO
     instead of using more specific todo keywords.
  3) todo state transition utilization.  Draw a nice
     graph-theoretical chart.
  4) space profiling, where we do something like the "du"
     command in Unix, counting the subtrees that are large
  5) linting, where we check for integrity of the outline
     tree, checking that all nodes have parents, aligning
     tags, enforcing indentation preferences, etc.

Myalgic encephalomyelitis causes death (Jason et al. 2006)
and severe suffering.  Conflicts of interest are destroying
research. What people "know" is wrong. Silence = death.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]