-How to deal with stuck / long-running jobs?
+Allow the output of a job to be listed. This is pretty easy since we
+know the tmpdir and the output file.
- - We should have instance/timeout settings that allows us to deal
- with jobs that run too long:
+For Whentools.once and Whentools.max:
- job "foo" timeout 1 hour ... # timeout the job after an hour
- job "bar" single instance ... # only one instance can run
- job "bar" kill previous ... # kill previous if still running
+ - Add a ~group parameter so we can run no more than "n" jobs in one
+ group.
-Should be possible to query jobs, running jobs, and kill (maybe even
-start) jobs.
+ - Add a ~kill_oldest (bool) parameter so we can kill the oldest jobs
+ and start the new job.
Implement native dynlink.