From the REPL:
(start)
To start the GUI:
It will watch joblib/-queue for changes and update itself.
Click generate 20 jobs to add 20 dummy jobs to the queue that tick down the seconds until they're done.
Click do 3 to start 3 jobs in parallel.
Click do all to start the remaining jobs in parallel.
Click clear jobs to remove the jobs from the queue.
Click generate 20 jobs to add another 20 dummy jobs to the queue.
Click do 3 until done to add a monitor that processes 3 jobs in parallel until all jobs are done.
Click generate 20 jobs and see 3 jobs taken in parallel again, until all jobs are done.
Click delete monitor to stop jobs being automatically processed.
From the CLI:
;; bring up the gui
(start)
;; add 20 dummy jobs to the queue that tick down the seconds until they're done
(run! joblib/create-job-add-to-queue! (gen-fns 20))
;; run *all* the dummy jobs
(joblib/start-jobs-in-queue!)
;; or, run just the first 3 jobs
(joblib/start-jobs-in-queue! 3)
;; or, create a monitor that watches the queue and ensures there are always 3 jobs running
(joblib/monitor! 3)
;; to stop the monitor from consuming jobs, keep a reference to it:
(def kill-monitor (joblib/monitor! 3))
;; and then call it when you're done:
(kill-monitor)
;; fetch results from the queue with:
(joblib/all-job-results)
;; or, fetch and remove the jobs from the queue with:
;; (this returns *all* of job information, not just job results)
(joblib/pop-all-jobs!)