>>Building an exact duplicate of a Saturn IB--assuming it were even possible--would be extremely wasteful<br /><br />I am not suggesting that. But it seems unlikely the Saturn I would have been built if an existing vehicle of adequate performance had been available, as is now the case.<br /><br /> />>Make it man-rated from the get go.<br /><br />This is just my opinion. I truly respect other viewpoints, and I could very easily be wrong. <br /><br />But I don't recall the Redstone, Atlas, or Titan II killing anyone, and the Soyuz is currently probably the most reliable launch vehicle in the world. All were originally designed as missiles. The best predictor of reliability is the number of launches a system has made, and launch vehicles that are used primarily for unmanned payloads are likely to have made more launches.<br /><br />I appreciate the intent of man-rating, but I think it gives us a false sense of security. It assumes that analysis can find all the possible failure modes and redundancy is the best way of protecting against them. In reality most catastrophic failures are unanticipated and can be eliminated completely by design changes once they become obvious. Man-rating, by focussing so pervasively on "fail-operational, fail-safe", has unfortunately made redundancy more important than reliability. Redundant man-rated O-rings didn't help Challenger. We now have four redundant LH2 MECO sensors in a system that was man-rated from the get-go and it still doesn't work.<br />