Were one unfortunate enough to collide with a chunk of FTL mass, this would result in immediate disruption of the Markov Bubble and immediate return to STL space, with possible catastrophic consequences, depending on one’s location.
The less energy you use to generate a Markov Bubble, the increasingly delicate the bubble becomes. Large gravitational hills, such as those around stars and planets, are more than strong enough to disrupt the bubble and dump you back into normal, subluminal space. This is what is known as the Markov Limit. With adequate computational powers, the limit can be lowered, but it cannot be removed entirely. Currently, Markov Drives cannot be activated in a gravitational field stronger than 1/100,000 g. This is why, in Sol, spaceships have to fly out to a distance equivalent to the radius of Jupiter’s orbit before they are able to go FTL (although if you’re actually near Jupiter, you’ll have to fly out even farther still).
As annoying as the Markov Limit is—no one likes having to sit through several more days of travel after weeks or months in cryo—it has actually proven to be a good thing. Because of it, no one can drop an FTL asteroid directly onto a city, or worse. Were there no Markov Limit, every spaceship would be far more of a potential threat than they already are and defense against surprise attacks would be basically impossible.
We are also fortunate that the viscoelasticity of spacetime precludes superpositional bombs. If a ship in FTL space flies over a mass in STL space that produces less than 1/100,000 g, and the ship returns to subluminal space at that precise moment, the ship and the mass will push each other apart with equal force, preventing either object from intersecting. If they did intersect, the resulting explosion would be on par with an antimatter detonation.
Once a spaceship has entered superluminal space, straight-line flight is usually the most practical choice. However, a limited amount of maneuverability is possible by carefully increasing the energy density on one side or another of the bubble. This will cause that side of the spaceship to slow, and thus the vessel as a whole to turn. But it is a gradual process and only suitable for small course corrections over long distances. Otherwise you risk destabilizing the bubble. For more substantial changes, it is better to drop back into subluminal space, reorient, and try again.
Any changes in heading that occur in FTL will be reflected upon returning to STL. Likewise any changes to total momentum/speed, with the degree of change being inversely proportional.
Technically it is possible for two ships in FTL to dock, but the practical difficulties of matching exact velocities, as well as the mathematics of merging Markov Bubbles, means that while it has been done with drones, no one—to our knowledge—has been crazy enough to try it with crewed ships.
Although a ship within a Markov Bubble can never directly observe its FTL surroundings, some level of sensory information is possible. By pulsing the bubble at the appropriate frequencies, FTL particles can be created on the outer surface of the membrane, and these can be used both as a form of radar as well as a signaling mechanism. With careful measurement, we can detect the return of the particles when they impinge upon the bubble, and this allows us to interact with superluminal space, albeit in a crude manner.
This is the same method by which FTL comms and sensors work. Both may be used far closer to a star or planet than one can maintain a Markov Bubble, but as with the bubble, there is a point at which the associated gravity hills become too steep for all but the slowest, most energetic FTL signals to climb.
Due to the protection of the bubble, a ship retains the inertial frame of reference it had prior to FTL, which means it does not experience the extreme time dilation that an exposed superluminal particle would. Nor does it experience any relativistic effects at all (the twins of the famous twin paradox will age at the same rate if one of them takes an FTL flight from Sol to Alpha Centauri and back).
This, of course, leads us to the question of causality.
Why, one might ask, doesn’t FTL travel allow for time travel, as all the equations for special relativity seem to indicate? The answer is that it doesn’t, and we know this because … it doesn’t.
Although that may seem facetious, the truth is that the debate remained unsettled until Robinson and the crew of the Daedalus made the first FTL flights. It took empirical experimentation to answer the question of time travel for certain, and it was only after the fact that the supporting math and physics were fully developed.
What was found was this: no matter how fast a superluminal voyage—no matter how many multiples of c your spaceship travels—you will never be able to return to your origin point before you left. Nor for that matter can you use FTL signals to send information into the past. Some amount of time will always elapse between departure and return.
How is this possible? If one is at all familiar with light cones and Lorentz transformations, it should be blindingly obvious that exceeding the speed of light results in being able to visit the past and kill your own grandfather (or something equally absurd).
Yet we cannot.
The key to understanding this lies in the fact that all three luminal realms belong to the same universe. Despite their seeming separation (as it appears from our normal, subluminal point of view), the three are part of a larger, cohesive whole. And while local violations of physical laws may appear to occur in certain circumstances, on a global scale, those laws are upheld. Conservation of energy and momentum, for example, are always maintained across the three luminal realms.
Adding to that, there is a certain amount of crossover. Gravitational distortions on one side of the luminal barrier will have a mirrored effect on the other. Thus, an object moving in subluminal space will leave an STL gravitational distortion in the equivalent FTL space. Waves from the distortion will propagate outward at c no matter what, but the movement of the gravitational center will be less than c. And the reverse is true for a superluminal gravitational mass, which would leave an FTL track of spacetime ripples through normal, subluminal space. (Of course, no such FTL tracks were detected prior to the invention of the Markov Drive, but that was a result of—in most cases—their extreme weakness and the distance of most superluminal matter from the main body of the Milky Way.)
Note: it’s important to remember that just as anything moving faster than c in subluminal space could theoretically be used to arrange a causality violation, so too could anything moving slower than c in superluminal space. In FTL, c is the minimum speed of information. Above that, relativity and non-simultaneity are maintained, no matter how fast you might be going.
Even without the existence of a Markov Drive, we now have a situation where natural phenomena seem to be violating the light-speed barrier on both sides of the spacetime membrane, but again, without inducing any causality violations.
The question returns: Why is that?
The answer is twofold.
One: no particle of real mass ever breaks the light-speed barrier in either the sub- or superluminal realm. If one did, we would see all of the paradoxes and causality violations predicted by traditional physics.
Two: just as TEQs form the basis for every subluminal particle, they also form the basis for every superluminal particle. As their name implies, TEQs are capable of existing in all three realms at once, and they are capable of moving as slow as the slowest STL particle and as fast as the fastest FTL particle—which is very fast indeed, limited only by the lower boundary of energy needed to maintain particle coherence, and even then, TEQs can move faster still given their Planck energy of 1.