Giulio Eulisse

@sawenzel I see there is a number of PRs which fail all with: ``` ## sw/BUILD/O2-latest/log 448/448 Test #404: o2sim_hepmc ..........................................................................***Timeout 400.02 sec ``` I assume this is some timeout...

Thank you for your contribution. Would you be so kind to squash your commits, please?

Okey, most / all of this code is actually now merged, in the form of a plugin. @Barthelemy the next step would be to try things out on the staging....

@costing @Barthelemy the multinode issue is back. Did anything change? ``` ## sw/BUILD/QualityControl-latest/log 27/39 Test #33: multinode_test ..........................***Failed 35.75 sec 97% tests passed, 1 tests failed out of 39 ```

@Barthelemy this is a CCDB issue, right? ``` ## sw/BUILD/QualityControl-latest/log 23/42 Test #25: testCheckWorkflow .......................***Failed 11.78 sec ```

Indeed we should check that `distance(begin, newEnd) + fake channels != distance(begin, pollOrder.end())`, no?

First of all, it's not clear to me how doing something in Subscribe() will avoid the state machine to advance. Properties are notified from a separate thread and the mutex...

@davidrohr as discussed. You might have to remove your callback, though.

What I think it's happening is the following: * A lock is acquired in shared memory to notify the events by the notification event * The processing dies of sudden...

That will mean that I need to check every time if a new region has been added, no?