The Martian is based on an appealing fantasy, but it's not a good basis for a manned space program. In my conversations with astronauts about human vs. robot roles in space, they convinced me that when it comes to distant places like Mars and the moons of Jupiter and Saturn, they'd prefer that robots do the advance work, dig shelters against cosmic rays, and turn up hazards long before the humans touch down. Quite likely, the first astronauts to arrive in the vicinity wouldn't land at all. Rather they'd handle the robots' mission control from orbit. But that's too wonky for fiction.
Lawrence in Arabia: Many good insights about T.E. Lawrence, the scheming British and French, the Turks, residents of Palestine, and the Saudi tribes. We hear much about what drove T.E. to such heights and such depths. But I found the side stories of other intriguers so prominently mentioned on the jacket -- presented in a parallel thriller style -- as weakly tied to the main story, because those people had little to do with Lawrence's exploits. Are they there for contrast, to show how much Lawrence did and how much they didn't do? I couldn't tell, but it was distracting. I did appreciate the occasional revisionist info correcting fictions in the David Lean motion picture.
Time and Again: A classic. A wonderful book, worth reading numerous times. Jack Finney is one of the few time-travel writers to realize that great narration and characters are more important than dropping time-travelers into the Turning Points of History. One of many examples of the latter cliche is The Final Countdown, where the writers dispatch a modern aircraft carrier through a "time storm," conveniently to arrive off Hawaii hours before the attack on Pearl Harbor. Here's the poster:
The Theory That Wouldn't Die: Reading this now: nonfiction about the origin of the Bayes Theorem and why it took more than two centuries to catch hold. Why does an old theory about using prior scenarios and posterior corrections matter? It's proven vital to good decision-making in uncertain and fuzzy situations. If the wreck of Malaysia Airlines MH370 is ever found -- no, make that when it's found -- the use of Bayesian methods is likely to share some of the credit ... as it did with the location of the remains of Air France Flight 447.
