Programmatic scheduling of quantum jobs - some leakage

[jarlah]

Hi

I until recently used add_job to upsert jobs, but I found quickly out that

[debug] Scheduling job for execution

was repeated for the amount of times I had edited the database row that triggers the add_job in the scheduler.

Therefore I have now added delete_job as a step, before I do add_job.

However internally add_job is actually deleting the job, so my question, is there something wrong with add_job that makes it leave something behind ? Maybe I missed some part of the doc that said that add_job can't be used as an upsert mechanism.

[jarlah]

I'm using 3.5 aka {:quantum, "~> 3.5"},