I'm a little surprised the consensus seems to be on using the direction to the main world from the sun as a 0 degree reference. And I'm not sure I understand the logic for dismissing using the galactic core, although using some significant star/pulsar males sense, especially if it can apply across a political entity (from the pov of setting navigational standards).
I like the idea of setting the system plane according to the primary stars equator. I assume its not too difficult to determine when first arriving in system?
Thinking on it, I guess ideally both your primary co-ordinates ought to be either system based or external/galactic based. So ether 0 degrees is the line from the sun to main world and the suns equator determines the system plane, or the galactic center/pick a pulsar is 0 degrees and the plane is the galactic plane.
I suggested to use the line form the main star to the primary world as reference (0 line) because it would be (IMHO) the easiest reference to pick when you enter a system, even though it rotates constantly arround the star.
Anyway, as all bodies in a system rotate arround the star, the position of any body will be changing at every moment, so I think no more complexity is added jsut because the 0 line rotates too (even it's a little less complex, as the main world will be in a constant (relative) position).
About the third dimension, you seem to put the 0 at the main star's equator (or at least so I understand), while I suggested the rotatory plain of the planets (which, AFAIK is more or less the same for all the planets, unless there's a captured one). I'm not sure if stars have also axial tilt in this sense...
If the plane for all the planets is not the same (or there are significant variation on it), I guess the main world should be used as reference too, for the same reasons given above.
The other consideration is whether in-system travel should use different reference points to jump travel.
Why would that be needed?
As I understand jump (I may well be wrong, off course), the jump points depend on gravitatory fields, and that's why they change all time and must be calculated for every jump. If so, and as the most used ones are the ones arround the main world (IMHO), this reference method would work quite well.
And I insist any reference method you use, all the positions in the system (system bodies, jump points, etc) will be constantly changing sue to the movement of all bodies in their orbits arround the star.
And a final thought, when measuring your location in degrees relative to the sun, while on approach to earth, to how many decimal places must you measure to get a +/- 10m3 accuracy?
I guess quite a lot...
But why do you need such accuracy? IMHO sensors would make quite less accuracy acceptable.
And even if needed (let's say jump points need it, and they are not detectable with sensors), I guess computers as represented in TU will be able to cope with as many decimals as needed.
EDIT: remember in my first post in this thread I also pointed the possibility the position could be expressed in radians, instead of degrees (as most TU uses SI units). If so, even more decimals would be used in any calculations, and most would iclude the number
pi on them.END EDIT