So here's the plan
Conquering space in two easy steps
Further to the discussion about Brian Wang's treatment of Orion and its offshoot, the Verne gun, if you look at the comments to my previous post, Adam Crowl suggests that peak acceleration for Brian's gun would be about 3700 gravities. He also suggests ways of reducing that, primarily by using a nuclear charge to energize hydrogen gas and have that push the ship. (I'm not sure that's the most efficient way to go, though, because the Orion design depends on the efficiency of energy transfer to the pusher plate and requires close proximity to the charge.)
In any case, this figure of 3700 g's suggests something: some things would be able to take it (like hardened electronics, tight rolls of thin-film solar cells, and liquids like water or rocket fuel) but others (like people and furniture) would not. In one of Brian's latest posts, he talks about the Mercury laser, which might make practical laser-initiated fusion happen. This piece makes me wonder what the total mass of the system minus the supporting building structure would be (because that bears on how practical it would be for fusion powered spacecraft) but also reminds me that laser launch systems have only been waiting for this one development to become practical.
So here's the plan: launch a few hundred thousand tonnes of rugged stuff using the Verne gun, and send up the rest a tonne at a time using a laser launch system. You can even run the laser launch system off renewables if you want to be green; and after the first few launches, you end up running it off beamed power from the first solar power sat you put up. True bootstrapping through hybrid launch technology.
Mag-Beam
Actually a decent Mag-Beam can be done for less in-space mass because a big ultra-light SPS array can provide the power directly instead of batteries. Using L'Garde inflatables for collectors the array can mass just 130 tons for a 300 Megawatt Mag-Beam power-supply.
As much as I like "Orion" I just don't like the idea of a drive needing massive amounts of bomblets for propulsion. The less we have to play around with nuclear explosives the better - though for mining on airless worlds they're pretty handy. External ignition of fusion pellets would be better and (relatively) non-weaponizable approach for pulse drives. Problem is that such ignition is hard to do! We know how to make fission bombs much better than we know how to ignite fusion without fission triggers.