Water from a comet
When my old employer the Jet Propulsion Laboratory slammed a 370 kilogram projectile into comet Tempel 1, it caused the release of at least 250,000,000 kilograms of water from the comet over a period of two weeks, according to the BBC. The peak rate occurred at 5 days, and it's a reasonably continuous flow. This suggests a variation on my ice rocket manufacturing technique: instead of going to all that work to extract, liquify, distill, and then freeze water from the frozen mudball, just make this kind of coma (something like a vapor geyser) and put the slowly rotating and shaded cylinder forms in the path of the coma. One probably has to tether the cylinders to the comet to keep them from blowing away, and I haven't worked out what the best distance from the comet surface is, or indeed if the deposition rate on the forms will be fast enough at a safe distance (deposition from gas is a different regime from the freezing spheres I proposed in 1992). Also the comet throws off carbon dioxide, methane, and other gases with the water, and the vapor carries some dust, so it's unknown whether the purity of the deposited ice will be sufficient to avoid overly rapid corrosion of the thermal ("steam") rocket. But the economics are potentially quite attractive so it's worth looking into. The basic economic advantage of rockets using native ice as the propellant source is that (once we've moved the ice to the starting orbits of our desired journeys) they save not only on the expensive launch of propellant from earth, but also on most of the remaining mass one must now launch out of earth's vast gravity well, namely propellant tanks. The BBC Tempel 1 story comes via Emergent Chaos.