Thomas James notes something that I didn't get around to noting yesterday--how limited in his thinking Stephen Hawking is:

If you're going to have to terraform even barren worlds with Earth-like parameters, how is that so much different from developing Mars-like planets as well? Why be so picky?

In fact, there just happens to be a Mars-like planet nearby, which wouldn't require anti-matter rockets or tens of thousands of years to reach...

And for that matter, there are plenty of asteroids and moons in the universe, not to mention the infinite possible variations on O'Neillian space settlements. Settling Earth-like planets isn't the only way to preserve the species.

Hear, hear. Planetary chauvinism happens to be a special interest of mine. And at the time of Hawking's statement, I too was struck by the fact that even an undisputed genius can be prone to it.

I think it relates to the fact that some things fit into our imaginations more easily than others. This influences science fiction which in turn influences what does and doesn't easily fit with the imaginations of the public at large.

People can be dumb, accidents happen, calamities are a fact of life. Cultures can loose vital technology - ask the Tasmanians who used to know how to make boats and sew but by the time Cook showed up had forgotten how.

If what you're after is species backup then having a place where you can breathe and breed without any technology at all would seem to be a good idea.

I'm all for having non-terrestrial habitats. But we need terrestrial worlds as well.

I suppose it would be foolishness to argue that one couldn't more likely survive a fall of civilization on a terraformed planet than on an O'Neill habitat. But I would make two points:

1. I think even a terraformed planet would require occasional maintenance to ensure survivability. Granted, on an O'Neill habitat, such maintenance cycles might be measured in decades while on a terraformed planet they might be several centuries. But still, I think it would be an error to conceptualize a terraformed planet as being as stable as the Earth.

2. It might surprise us how maintenance-free one might engineer an orbital habitat to be. There's every reason to expect a rotating habitat to have a "de-spun" section. But in a pinch you could dispense with this, and the entire habitat could rotate as a monolithic unit. For heat rejection it was expected that we would avoid the complications of compression/expansion and settle for simple air circulation in heat radiating panels. But if we wanted to eliminate even the simple electric motors driving blowers, perhaps some totally passive heat rejection system could be implemented. You finally get down to the motors and timers involved in tilting the mirrors for the day/night cycle. Might those be engineered to be reliable on a time frame of a few centuries?

This influences science fiction which in turn influences what does and doesn't easily fit with the imaginations of the public at large.

Perhaps we need to pressure hollywood into an Iain Banks or Peter Hamilton movie deal?

I suppose it would be foolishness to argue that one couldn't more likely survive a fall of civilization on a terraformed planet than on an O'Neill habitat.

Until we try it a few times we're talking moonbeams - we just don't know.

What I'm thinking is less being able to survive for a few centuries than a place you could live in for a few thousand or ten thousand years. Talking about knocked back to animal skins and mud huts and having to rediscover everything from scratch.

One problem with terraforming Mars - no magnetosphere (source), thus no natural protection from solar flares and cosmic radiation. IIRC, solar flares erode the Martian atmosphere, which means that some solare flare protection (an artificial magnetosphere?) would have to be in place before atmospheric terraforming could even begin.

How hard would it be to terraform the Sahara?

One problem with terraforming Mars - no magnetosphere (source),

This is actually one of the easier problems to solve. Run superconducting cables around lattitude lines and charge them up with persistent currents. This will be a macroengineering project, but a trivial one compared to the actual terraforming of the atmosphere and biosphere. You don't need a field as strong as Earth's to exclude most cosmic radiation, so the energy stored will be surprisingly small, equivalent to a year's power output from a moderate sized industrial turbine.

IRC, solar flares erode the Martian atmosphere, which means that some solare flare protection (an artificial magnetosphere?) would have to be in place before atmospheric terraforming could even begin.

Huh? If solar flares eroded the atmosphere over the time required to terraform, the planet would have no atmosphere at all right now, since it's been around for millions of times longer than that.

Relevant to the topic of atmo terraforming and my faulty memory on flares vs. Mars is the Wikipedia article on the atmosphere of Venus:

On Venus however, it has been theorized that the much weaker weak magnetic field (only 0.000015 times that of Earth) has caused lighter gases to be knocked out of the atmosphere, leaving only carbon dioxide and small traces of other elements and molecules.

So, if theory holds, the artificial magnetosphere would still be needed prior to atmo terraforming, but not because of solar flares.

Atmo terraforming would thus require two separate technologies. One would extract the carbon from CO2 on carbon-dioxide-rich planets. Another would introduce oxygen into an atmosphere where CO2 is not abundant. (By transforming atmospheric nitrogen and/or other heavy gases into oxygen?)

Any reason not to terraform Venus? I suppose we would have to destroy much of its atmo to get the atmospheric pressure right...

Other than the fact that it might be harder to cool a planet down than to warm one up, I see no reason not to try to terraform Venus.

So, if theory holds, the artificial magnetosphere would still be needed prior to atmo terraforming, but not because of solar flares.

No, this remains a non sequitur. The process on Venus has operated over billions of years. This does not imply it would be significant over a period of thousands of years. Numbers matter, Alan.