worklist naming change to be more descriptive

[petersilva]

So... in worklists, there are the following sublists:

• incoming ... messages from gather or retry...
• ok ... processed successfully.
• failed ... tried to do the work, it failed, will be queued to retry later.
• rejected... will not attempt transfer/work. reasons:
  • accept/reject
  • mtime (too new, too old.)
  • duplicate (checksum match)
  • metadata mismatch (mtime of message older than what we already have.)

---

So when we attempt to retrieve a message... if it fails, whether it goes to failed or rejected depends on why it failed... for a 404... it would go to rejected... It isn't really rejected, it's actually failed in a way that retry will not help with, so we should not be putting it on retry queue.

• I think failed would better be called... retry_later to better describe the purpose of putting messages there.

• I think rejected is better described as discard

Making this change, is a breaking API change, since all plugins need to be modified, as they use the worklists.

It seems like an improvement in readability/ease of understanding... but... dunno... maybe not worth it.