space stations vs terresterial colonies

>Maybe carbon nanotube space elevators will save us from this grim
calculus, but that's still very iffy.
> Hang out at the space elevators listserve for awhile if you don't
believe
> me, and read some of the technical hurdles still to be overcome.

Is anyone here discussing other options? I like the idea of a mass
driver (electromagnetic launch system, linear accelerator, whatever
you want to call it) accelerating a rocket to the speed of a surface-
to-LEO or surface-to-wherever transfer orbit. It would require a
tube to an altitude where the atmosphere is thin enough not to burn
up the spacecraft, but that seems like a much smaller project than a
skyhook for most of the benefit. Payloads going through the tube
would push the air out; if it was down for a while and filled up,
dummy payloads would be put through to sweep out the air. It would
need to be effectively stiff enough not to have changing winds blow
it into the wrong curvature and smash payloads into the sides, but
that could be achieved by having balloons sufficient to support
twice the weight and tethers holding it down from multiple angles.

> Given these limitations, where will humans first live? Where the
resource
> mining and manufacturing is, so that they don't have to transport
materials
> long distances to where they want to build.

Well, not *distance* per se, but delta-vee.

Distance is important for light-speed communication: you can't use
remote control if there's a ten-minute light-speed lag between
twitching your joystick and seeing whether the robotic hand went
where you intended it to. For on-site operations at an asteroid,
you need to use autonomous machinery that makes its own real-time
decisions. The control has to be limited to having it send progress
reports and having you update its programs occasionally. For
operations in LEO, on the other hand, it takes less time than the
lag on your international phone calls made via satellite.

> Imagine if you could build a facility that could deliver to low-
earth orbit
> bottled O2 and N2
...
> If you could put a machine on an asteroid that could
> deliver 100 bottles a year back to LEO (it would obviously have to
make the
> bottles too, and some sort of transport mechanism),

I don't want to purify the gases and manufacture the bottles in the
asteroid's original orbit. When the mining spacecraft reaches the
asteroid, I want to start separating out *low*-value components of
the asteroid, and then use them as reaction mass to bring the rest
of the asteroid into earth orbit. It would arrive with perhaps half
the material having been through the first separation step, and the
other half completely raw materials. Then you could send up new
equipment if necessary any time you hit a snag in processing, or you
could send humans up to do a few of the trickier operations, instead
of having to wait two years for a launch window to send another
machine.

> that's an annual income
> of $3,000,000 a year. If you could build it to last for 50 years,
it would
> generate $150 million in income over that time. So you need to
make
> something that you can put up for less than $150 million, that
will last for
> 50 years untended, and you're making a profit.

Time is money. If the venture is low-risk enough to get by with a
10% interest rate, an annual income of $3M/year is worth $30M/year
if it goes on forever. (The $3M is what you would get as a 10%
return each year on a $30M investment.)