From #123.
Allow hardware devices that could run on the rails, and trigger an interrupt on the DCPU. For example, you could have a "Sphere-of-Influence-omatic" that would raise an interrupt when a vessel changes SOI. A DCPU could register for this interrupt, halt itself, which would then allow it to go on the rails, then periodically we could check for SOI changes. When we detect one then we could alert the user "Hey, do you want to switch to this vessel so its DCPU can do its thing?" If the user agrees we change to the vessel, making it active and taking it off the rails, raise the interrupt, and the DPCU continues executing merrily. This would be a post-1.0 thing.
From #123.