While I love the stories of C.J. Cherryh (Merchanter's Luck, Chanur's Legacy, etc.), the designs she used could really be improved. On the other hand, she also got quite a lot right.
Shuttling between a planet and orbit requires entirely different type of vehicle than interplanetary or interstellar travel. Not only do they have to be aerodynamic, but they need high thrust more than high efficiency. Consequently, the station orbit should be low to minimize the travel time and delta V for the shuttles.
It should be noted that it can take several days for resupply ships to rendezvous with the ISS. The reason for this is that one must launch when the orbital plane for the station passes your location (which happens twice a day). The station could easily be on the opposite side of the Earth at this time, so one must slowly catch up using a lower orbit.
If the station had an equatorial orbit, one could launch so as to minimize travel time (about once every 90 minutes), but only from the equator. For other locations one would have to fly to the equator first, or burn a lot of fuel changing the orbital plane later.
Attempting to dock a ship to or near the edge of a spinning disk is a recipe for disaster. The only safe place to rendezvous is on the axis of rotation. On the other hand, undocking from the edge is even safer than from the center, since the ship will simply fall away (along a tangent).
One solution would be to place docking arms near the axis, adding an extra margin of distance and shock absorption. Specifically, one arm would place a docking target along the axis of rotation at some distance from the station bulkhead. After docking, the arm would then transfer the ship away from the center to its berth location. Multiple arms are needed to access all of the berth locations and to allow multiple ships to dock quickly. For purposes of cargo transfer, it would be better to dock in a low G radius rather than all the way to the outside edge.
To avoid thrusting directly at the station, ships should take spiral trajectories, at least for the final approach. Space is big, so establishing regular lanes is unnecessary. It should be easier to simply allow only one ship at a time to approach within a certain radius, and assign docking times in advance so that ship captains can time their approaches.
Ship transponders broadcasting basic navigation data would allow ships to avoid each other and especially avoid crossing rocket exhaust plumes. Particularly useful would be navigation "buoys" to function like GPS satellites.
Ship berths are more likely to resemble airport terminals than naval piers. Each would have quarantine/customs/security/breach bariers, plus temporary cargo holding facitilities. Cargo transport is more likely to be an automated system than guys driving around in trucks.
For permanent habitation, avoiding radiation will be a major problem. It would be a good idea to line up the axis of rotation to point at the sun and place the thickest shielding there. A large parabolic reflector and some sort of thermal power generator would be more effective than arrays of solar cells.
It should be noted that the station will need to continually adjust its spin rate and precess the axis alignment toward the sun.
So with one end used for docking and the other for solar power/shielding, the only way to expand is to lengthen the cylinder. The station would therefore consist of disks separated by air tight walls. To expand, one would close off two disks, separate them, and then build a new disk in the space between.
One could add additional barriers for fire and blowout prevention, but those are optional. Each section should be capable of independent life support with minimal connections between them.
The outermost layer would be used for work and recreation, with housing immediately above, so that people could spend as much time in high G as possible (for health). The innermost sections would be for storage, with various forms of farming between.
It takes a
The best way to provide a 1G environment in a space ship is to simply accelerate or decelerate constantly at 1G. In fact, one would need to accelerate at least that hard just to get out of the solar system in a reasonable period of time. The only time spinning would be useful would be while drifting through an uninhabited system (which is a peculiarity of Cherryh's form of hyperspace drive). Even then, one would probably need to change directions (as much as 120 degrees).
The standard cargo container has revolutionized sea and rail transportation, with limited penetration into trucking. Air cargo uses smaller containers with no general standard. The ideal size for a standard space cargo container may take a long time to establish, but the "can" used by Cherryh is a good guess.
An effective way to store cans in a ship would be to arrange them on circular tracks (think lazy Susan) just inside the bulkhead, with several elevators to move individual cans from one track to another. One or two tracks would connect to cargo doors for loading and unloading. Some sort of inspection access would also be needed.
Larger cargos would have to be attached externally, using either a different standard or custom container. The ship would then need an array of external hardpoints for attachment.