changemachine changelog

0.3.2

Made it always ok to process, will address ability to pause the machine separately to the change notifiers at a later date. For now stability of processing more important.

0.3

Added machine registry and repl
Reworked fail event to pass through error

0.2.2

Rework "at capacity" detection logic
Refine machine.close logic to clean up neuron job queue
Only allow one machine with one id active at a time
Ensure that if a notifier is created for the machine, then it is paused if the machine is in a paused state.

0.2.1

Add the leaveOn member for a machine, by default leaveOn.fail is false

0.2.0

Bump version to reflect addition of new methods in 0.1.x versions

0.1.7

Added pause and resume methods and events to the machine for flow control help

0.1.6

calling done on an item with an error in the options, now maps to a fail
added a close method to a machine which closes the underlying changemate notifier
changed JsonStore to use fs.writeFileSync as in particular cases callback was never firing for fs.writeFile

0.1.5

Revisions to checkpointing logic (enable fails to checkpoint)
Added checkpointOn object to machine for toggling checkpoint on done, fail, etc

0.1.4

Made item methods chainable
State now referenced as checkpoint (it's more accurate)

0.1.3

Added basic state storage (JsonStore)

0.1.2

Added error message for item.fail call
Parameter passthrough for item done and fail events

0.1.1

Reworked items and queuing to make manual item / machine interaction simpler.
Added status queues in the machine for watching item state
Reworked newListener implementation to be more efficient.

0.1.0

Initial version