Separate declaration from initialization and add variable comments.

I don't understand how could I not have it there for so long.

We want to choose browsrt with bmo-sensible now, so sensible.ini holds
this and many more similar lists.

This will ensure nagging is predictable during session: on first
opportunity (which is normally new terminal window) `tasknag` will nag
about everything, then according to preset intervals.

I keeo forgetting to say "modify" and even `m` alias did not help.  This
maybe...

...yep, we're getting a bit naggy...

This likely breaks hostname setting when accessing via SSH, but:

 *  it's first version that is at least somehow readable,

 *  and also WE *MAY* have fixed the long-standing PITA escaping
    errors that we used to get in urxvt if BASH_COMMAND had backslashes
    in it!

 *  Remove excess comments and optional syntax (eg. EOL `;`),

 *  use non-conflict function names,

 *  simplify PS1 generation,

 *  close PROMPT_COMMAND setup into function,

This used to be re-set in host-specific bashrc, but we already rely on
so many things installed anyway...

This reverts commit 55a4088d807c5ca0552e89c3240ec1d3c88842f6.

Turns out this makes it too hard to explore and learn the vast potentials
of taskwarrior CLI.

At least once I ran into problems, not realizing that full / was the
cause.

Not funny.

AFK and GONE have dedicated key bindings, so they don't need to be
first.  BACK is never meant to be called manually anyway.

UNDOCKING and ZLEEPING may often be needed in hurry, though, DOCKING
normally comes soon after them.

Since task IDs are dynamic in taskwarrior, commands like `task 4 delete`
and `task 4 done` can be destructive with hard-to-trace damage if
inadverently repeated.

Better to avoid that, even at cost of not remembering anything.

(Note that `task` without arguments is still remembered; there's no harm
in repeating that, and it still can be convenient eg. w/ ^L).

Script not (yet) released, but it's s simple taskwarrior wrapper to
remind about unfinished and overdue tasks once per set period.

xclose just closes window instead of killing its owner, which would tear
down all this owner's windows along.

Uses different utilities with a different set of drawbacks, though:

 *  wmctrl(1) needs to be installed.

 *  [slop][1] can be optionally installed from source but can be unreliable
    when selecting some windows, e.g. urxvt.

 *  If slop is missing, xwininfo is used, which, unlike xkill(1), has no
    way of cancelling the user action (r-click selects a window just the
    same as l-click).  We work around it using timeout(1), which is
    rather inconvenient, though.

But all in all, we're still better off than with xkill(1).

  [1]: https://github.com/naelstrof/slop

This makes a git compatibility warning go away.

Don't let all the bmo processes hang

http://unix.stackexchange.com/a/18443/9365

Another proof that 640KB won't be enough for everybody.

This should enable reading upper parts of window even while the output
is still being produced (tail -f).

IOW, the auto-scroll is disabled after scrolling back manually using
S-PageUp or S-PageDown, and re-enabled on another keypress such as
typing when a prompt is expected at the bottom.

Hey, those numbers are *rough* estimates...

Turns out low_threshold is in minutes, not in percent, by default.
Needless to say, time estimates are grossly unreliable here and 10
minutes left is too late anyway.