When you create a watcher, you’ll notice it’s configuration file,
in the watcher folder:
$ tree /home/vanessa/.watchme/ /home/vanessa/.watchme/ └── watcher └── watchme.cfg
and what you can’t see is that there is a Git repo (a .git folder) in the watcher directory too.
By default, the watcher simply has a status, active or not.
[watcher] active = false
When you first create a watcher, it will not be active. When you activate a watcher,
this coincides with generating a cron job to run the command for it, and the cron
job will run a watchme command to read this configuration file. This means that you
can change the active status at any time that you need (manually or via
watchme activate <name> and the cron job will run and essentially do nothing.
You can optionally protect a watcher, meaning that you can’t delete it. Protected means that you can’t delete the watcher, but you can edit the tasks.
[watcher] active = true protected = "on"
If the parameter is missing or defined with value “off”, both coincide with not protected.
If you want to freeze the configuration entirely, you should freeze it. This means that you cannot delete the watcher folder, or edit anything in the configuration file (other than the frozen status).
[watcher] active = true frozen = "on"
If the parameter is missing or defined with value “off”, both coincide with not frozen.
Each action done by a watcher is called a task. A task is identified by starting with “task” in the configuration:
[task-reddit] url = https://www.reddit.com/r/hpc active = false
Akin to the watcher itself, each task can also be active or not. The variables within the task will vary based on the kind of watcher. For example, the above is from the urls watcher type, since it will watch web pages for changes.
This code is licensed under the Affero GPL, version 3.0 or later LICENSE. The SIF Header format is licesed by Sylabs.