Running ‘alive’ is easy. Simply type ‘alive’ at the shell prompt followed by RET. You can use the shell’s i/o redirection facilities to filter/suppress the output, and its job control facilities to have ‘alive’ execute in the background. For example, with GNU Bash (see GNU Software):
$ alive >/dev/null &
Two other modes of operation are available:
$ alive --version $ alive --help
That’s it! Quite boring, no? The real fun lies in playing with the configuration, either before starting ‘alive’, or while it runs. See Configuration.
Additionally, ‘alive’ responds to certain signals in a more or less well-defined manner. To send it a signal, first note its pid (process id), displayed on the first line output:
$ alive alive: 2012-08-11 16:24:26 restart (pid 9731) [...]
The pid in this example is
9731. Next, use the command-line
$ kill -INT 9731
or Emacs command
M-x signal-process RET 9731 RET 2 RET
specifying this pid and the signal to send. All unhandled signals (not described in the following table) will cause ‘alive’ to die unceremoniously. To see a list, try ‘kill -l’. Handled signals, on the other hand, elicit an acknowledgement on stdout, e.g., ‘received signal 2 (SIGINT)’.
Interrupt the current
sleep(2). This is useful if you change the
configuration and don’t want to wait for the next iteration.
Restart. This is useful if you forget the pid.
Shut down, i.e., display ‘exiting’ and exit successfully.