plainbox session restoring doesn't restore generated jobs if run crashes while running a generated job

[Daniel Manrique]

- Modify plainbox/stubbox/jobs/multilevel.txt.in, and add a "sleep 20" in multilevel_1 command (to give us time to kill plainbox)
- Run as follows:

plainbox -c stub run -w plainbox/stubbox/data/whitelists/stub1.whitelist

- When you hit multilevel_1 and it's waiting, open another terminal window and kill plainbox (I looked at the pid on ps wwuxa |grep plainbox and killed -9 by specific pid)

- Rerun plainbox the same way

- Select "Yes" to restore the session

This result shows multilevel_1 and multilevel_2 disappeared:

===============================[ Analyzing Jobs ]===============================
Do you want to resume the previous session [Y/n]? y
Estimated duration cannot be determined for automated jobs.
Estimated duration cannot be determined for manual jobs.
==============================[ Running All Jobs ]==============================
==================================[ Results ]===================================
stub/dependency/bad: not-supported
stub/false: fail
stub/multilevel: pass
stub/requirement/bad: not-supported
stub/requirement/good: pass
stub/true: pass
stub_package: pass

Whereas the output should be closer to this (this is letting multilevel_1 complete on a non-restored session run):

==================================[ Results ]===================================
stub/dependency/bad: not-supported
stub/false: fail
stub/multilevel: pass
stub/multilevel_1: pass
stub/multilevel_2: pass
stub/requirement/bad: not-supported
stub/requirement/good: pass
stub/true: pass
stub_package: pass

[Zygmunt Krynicki (zyga)]

I'm marking as incomplete. I don't think it's happening anymore but I haven't researched this thoroughly.

[Daniel Manrique (roadmr)]

OK, I retried the repro case and it's not happening anymore. Marking as Fix Released :D