Current thinking is a small/inflatable buried structure with absolute minimal facilities. An inflatable 4 metre diameter tube roughly 6 metre long, laying on its side, buried 2 meters, with two meters of regolith overburden. Say 3-5 bunks on a wall, a couple desks, a RV style shower/head and a small sink. Maybe if we are really lucky a place to heat food. There will probably be three airlocks, one outer lock leading to a suit room, a second lock between the suit room and the living quarters to keep dust down, and a third as an emergency exit. There will be a limited amount of water recycling, and until we get the in-situ resource manufacturing down, most supplies will need to be brought it. Of course power will be supplied by solar panels over the overburden, (A little more protection) and perhaps as a sun shade over some of the external equipment. Power storage for overnight is still in the 'air'. (Sorry)
We'll get a smaller image up soon.
(10/28/08) If we are prepared to accept a 2% increase in circumference for the inflatable outpost, we may be able to expand available floorspace by 30%.
A true cylinder 4 meters (12.8 ft)in diameter, constrained to a 2.5 meter (8.0 ft)floor-to-ceiling height, will have a floor approximately 3.8 meters(12.16 ft)wide. The circumference of a 4-meter diameter cylinder is 12.56 meters (40.19 ft).
A half-cylinder, or quonset, shape 5 meters in width will have a circumference of 12.85 meters - a 2.3% increase - yielding a floor area 1.2 meters wider - an increase of 30%. The floor-to-cieling height remains the same at 2.5 meters.
Actual curvature of the walls is increased yielding a slightly greater distance from the centerline for a quonset configuration based on these dimensions.
Given that floorspace and overall volume are always at a premium, I would recommend going with the quonset style for an inflatable. There would be the same number of seams (potential leakage points) in its construction as the cylinder and only a somewwhat greater complexity in construction, but actually this trades one complexity (installing a floor in an inflatable cylinder) for another (installing two 'D'-shaped endcaps to the ends of a fabric with non-uniform dimensions)for a net balance in effort. I'm drawing up an internal arrangement concept for this design.