discard_past doesn't seem to be working for "every" jobs with max_work_threads

[rscarborough1996]

Hello,

When an "every" job is delayed due to max_work_threads limit being reached, it will run multiple times when it finally gets the chance, which I would expect discard_past to prevent.

Example:

require 'rufus-scheduler'

scheduler = Rufus::Scheduler.new(max_work_threads: 1, discard_past: true)


scheduler.in '1s' do
  sleep 5
  puts "job finished #{Time.now}"
end

scheduler.every '1s' do
  puts "tick #{Time.now}"
end

scheduler.join

The above code will produce something like this:

job finished 2023-08-24 17:22:14 -0400
tick 2023-08-24 17:22:14 -0400
tick 2023-08-24 17:22:14 -0400
tick 2023-08-24 17:22:14 -0400
tick 2023-08-24 17:22:14 -0400
tick 2023-08-24 17:22:14 -0400
tick 2023-08-24 17:22:15 -0400
tick 2023-08-24 17:22:17 -0400
tick 2023-08-24 17:22:18 -0400
...

As you can see, multiple "ticks" occur at once.
Is there a work around, or perhaps I am missing something?

Thanks

System:
ruby 3.0.2
rufus-scheduler 3.9.1
running in wsl2 Ubuntu

[jmettraux]

I did not forget this one. I think it needs some rework. Give me please a bit more time.