By adding the "kind" to the base resource, it is still identifiable even
when the resource specific elements are gone in calls that are only
defined in the base resource. This is also more logical for building
resources!
This also switches resources to use an Init() method. This will be
useful for when resources have more complex initialization to do.
This allows for resources to automatically add necessary edges to the
graph so that the event system doesn't have to work overtime due to
sub-optimal execution order.
I added a regression to the file resource. This was caused by two
different bugs that I added when I switched the API to use checkapply. I
would have caught these issues, except my test cases *also* had a bug! I
think I've fixed all three issues now.
Lastly, when running on travis, the tests behave very differently! Some
of the tests actually fail, and it's not clear why. As a result, we had
to disable them! I guess you get what you pay for.
This used to be GetType(), but since now things are "resources", we want
to know what "kind" they are, since asking what "type" they are is
confusing, and makes less logical sense than "Kind".
This simplifies the API, and reduces code duplication for most
resources. It became obvious when writing the pkg resource, that the two
operations should really be one. Hopefully this will last! Comments
welcome.
Naming the resources "type" was a stupid mistake, and is a huge source
of confusion when also talking about real types. Fix this before it gets
out of hand.
* Fixup graph state readability
* Rename original SetState() to SetConvergedState() and friends...
* Add type state management for proper BackPoke() commands...
* Add better DEBUG logging
This is an important optimization that prevents running a BackPoke on a
parent which is in the process of running and will most certainly poke
the caller back in a moment. This avoids unnecessary roundtrips.
Unfortunately, there are still other algorithms required so that races
can't cause the graph to run for longer than necessary.
* Fix Process() object calling
* Add PokeParent() to poke upwards
* Break linear exec chains :(
This was the issue where in a graph f1 -> f2, if you were to rm f2 &&
cat f2, then f2 would not come back because we didn't poke upwards to
refresh the timestamp. Unfortunately this adds another bug which we
solve in a later patch.
* Add exec type
* Switch erroneous use of fmt to log instead
* Check for edge existence for safety before using
* Avoid recalling etcd channel maker
* Clean up logging output
