wastedyears is a tool to track the time you spend getting stuff done. Typically this is for a job that you do on a computer. It lets you keep track of the tasks you work on, and the number of times you switch between those tasks.

Although wastedyears is named after an Iron Maiden song, its official theme song is Pink Floyd's 'Time'. Listen to the lyrics to find out why.

For example, after entering a couple of tasks into wastedyears, your day might look like

08:57:17 .. 09:03:42 check email
09:03:55 .. 09:06:15 coffee
09:06:27 .. 10:15:22 fix bug #321
10:15:42 .. 10:23:17 check email

If this strikes you as insanely detailed, and the sort of thing that you would never in a million years care about, then wastedyears might not be the right tool for you. However, it is carefully optimized to minimize the time spent tracking your time, in order to maximize the time you spend doing useful productive work. (Or checking email and getting coffee, as the case may be.)

There are a couple of ways to enter a task into wastedyears: command-line and GUI.

At the command line, you can run

wy t check email

to end the previous task and start the next one.

Or you can simply run

wy t

to enter a task into a text editor. The editor will actually show some recent tasks, so you can copy/paste frequently used items:

# wy task editor; enter new task here:


# recent tasks from yesterday:
amusing email to Andy
boring email to Bob
check email

# recent tasks from today:
check email
coffee
fix bug #321

Once you close your editor, wastedyears adds the first non-comment line to its database as a new task. Your previous task finishes when you launch the editor, and the new one starts when you close it. That's how you get those gaps in the record -- 10 seconds here, 15 seconds there, the overhead of recording your tasks.

Finally, you can run

wy gui

to open a small graphical interface that is designed for a quick get-in-get-out experience. It's the same basic idea as running a text editor, but more specialized.

wastedyears stores your tasks in a relational database. Currently only SQLite is supported. The database is kept in $XDG_DATA_HOME/wastedyears/wastedyears.sqlite3.

There are two tables in the database: tasks and words. The tasks table is fairly obvious:

task_id  start_ts  end_ts    updated_ts  description
-------  --------  --------- ----------  ----------------------
      1  08:57:17  09:03:42  08:57:17    check email
      2  09:03:55  09:06:15  09:03:55    coffee
      3  09:06:27  10:15:22  09:06:27    fix bug #321
      4  10:15:42  10:23:17  10:15:42    check email

(All of the *_ts columns are actually date-time columns; I'm showing only the times here to keep things simple.)

The words table, which is derived from tasks, is where things get more fun:

word_id  elapsed  word
-------  -------  -------------------
      1      840  check
      2      840  email
      3      140  coffee
      4     4135  fix
      5     4135  bug
      6     4135  #321

elapsed is the total number of seconds spent on tasks that include that word.

Now you can finally prove to your boss that you spend more time fixing bugs thatn getting coffee. And, after a while, you'll know which bugs took more time.

Not yet solved: do you spend more time checking email or sending email? I think that will require another derived table to record how strongly words are correlated with other words. TBD!

There are a couple of command-line tools to list what's in the database:

wy ls-tasks

wy ls-words

By default, these list in a simple plain-text format. Use --json to dump the data in JSON.

Once you have built up a few days' (or weeks') worth of tasks, it's interesting to dig into the data a bit and see what you have been doing. If you're comfortable with SQL, you can of course do this entirely manually:

sqlite3 $XDG_DATA_HOME/wastedyears/wastedyears.sqlite3

So far, there is only one built-in analytics command:

wy rank-words

to list words from "most elapsed time" to least.