Other people musing... about colonies.

http://www.l5development.com/d_space_sys/d_colony/b_Design/DesignTheory.php

Here, volume, mass, gravity, area, etc. calculations were made in planning a colony ship, using sophisticated math (or, if not as sophisticated as it appears, then a lot of arithmetic. Compared to other models it may or may not be considered big, but the estimate for materials is 1.6 million metres cubed of materials, enclosing 46 million cubic meters of volume. Now, of course, the environment for structures is very different in space than on earth. First, collapse works very differently; it could usually only be crushed by its own mass, which would require some serious size and/or rotation. It could also risk obliteration or massive loss of course if another body's gravitational pull was great enough to affect our structure. Satellites can ignore many of the things mentioned here, because they're small, orbiting slowly and relatively safely around the earth, and they don't usually have to survive impact with a planet (however gentle). A colonial ship would not only have to brave all the bits of debris that floated around space at very high speeds, it would have to get its human cargo off onto another planet safely. If we can't just teleport or shoot people off the craft, then the craft likely has to go down too. First, if it can survive its incredible bulk being launched off of the earth (if possible, an assembly station off of Earth), then it has to survive entry into another planet too. If each cubic meter of material weighs an average of 15kg (likely more), and then the craft loses, say, 1million tons of material (propellant and small amounts of unreusable waste?), then that's 15million tons of stuff that has to survive planetfall. Now, even if said planet has very low G, we're trying to get 15 million tons in a 46 million cubic metre thing to unload cargo into another major gravity well. Good luck working those physics (not too much sarcasm).

The technology for a self-sustaining environment of such, to me, doesn't seem incredibly far off, if we can't do it already. Small systems of plants and animals, in a jar, with sunlight and usually in water, have worked. Making it spin around in space and resist the forces it will create (rotating decks big enough to produce around a G) is a challenge. Part of the challenge here is getting incredibly strong materials with various properties to suit the needs of craft and passengers. Another dead end question that only continued research can come up with a real answer to.