Limitations / TODOs
- only allows one worker at a time due to hard-coded worker name
- only allows worker to read from one queue
- ttl & heartbeat not implemented -- longer running jobs won't work well
- need to support retry for certain error types / messages
- stalled jobs might not be handled right? (worst case is user has to go under "stalled" and timeout the jobs manually)
- might have to check how i handle redis connection/TCP-level errors -- may cause unexpected behavior (but qless is safe so jobs shouldn't be dropped on the floor)
- Limitations not to be fixed this sprint:
- lots of other unimplemented functionality, such as recurring jobs and tracked jobs
- requires node >= 4 -- should be avoidable with Babel -- maybe automatically with npm magic?
- need to look through Python and Ruby classes and see what functionlity
we have missed for the following classes:
- Job
- Jobs (ClientJobs)
- Queue
- Queues (ClientQueues)
- Client
- check if there are any other classes