Programmatically adding new EventListeners to a Configuration is straight-forward.

However, adding/removing event listeners to a running SessionFactory is semi-sort-of-possible-but-not-recommended. Look at org.hibernate.impl.SessionFactoryImpl. Notice that most member variables are final – including the EventListeners and Interceptor variables.

I would take this as a sign that Hibernate developers regard (rightly so) that such changes to the listener list as problematic. Especially since the actual core behavior of persistence is encapsulated in an EventListener! Suddenly for a period of time no data is being saved could be a possible race condition!

Internally, we have an EventListener implementation that delegates to a list of EventListeners. This allows the dynamic add/remove – however, we are only doing this because of some configuration oddities.

In normal practice I really would question if this is. Have you considered creating a new SessionFactory with the new EventListeners and Interceptors? Doing so is a more likely to be clean and avoid “interesting” race conditions.

Also why are you trying to do this modification at runtime?

Thanks for stopping by.

Your issues are:

1) turnout speeds require that a train slowdown to exist the main track to the local loop.
2) Right of Way widths with 4 tracks.

Turnout Speed limitations
———————
You are correct that a train will need to slow down because of limitations on turnout speeds. However, this point is meaningless for these reasons:

* The speed through the east bay ( just like along the peninsula if the Pacheco Pass route is used ) will likely not be at maximum speed.
* Even if the train is running flat out. Based on my back-of-envelope calculations slowing from 220mph to 140mph should only require that the following train be 5 minutes behind (instead of 3 minutes behind). This is a minor scheduling issue and does not impact the following train’s speed.

Constrained right-of-way
———————-
This is a red herring. The same “problem” only worse exists along the Caltrain ROW. But this does not seem to deter the California High-Speed Rail “authority” from deciding to drive the train through some very expensive real-estate. Most property bordering the train tends to be commercial anyhow.

In any event, 4 tracks requires only 80ft Right-of-Way, which is considerably narrower than many roads.

—————————————

The rest of your points fall into the category of interesting red herrings that do not pertain to the question at hand.

* “SNCF complain that pressure from local communities for more services to stop at their town, impedes them from running more efficient TGV services.” — So more people want the train to stop in their town… how is this related to the question at hand which is does the existence of a train station impact the speed of a train that is not stopping.

* “Reducing line capacity” is only meaningful if the traffic numbers warrant more trains. If the demand does not warrant more trains then the “reduction in capacity” is meaningless.

* Madrid – to – Barcelona observations: Interesting but meaningless. Missing data: When ? What time of day? Weekday? Weekend?

—————————–

My point is and continues to be: a freeway (or a rail system) can be engineered for higher speeds than possible for the current vehicles. China in fact is building their High-Speed Rail System to support trains running at 250mph anticipating that those speeds will soon be possible.