Why should I try all that hard when Drezner does such a great job? See his blog on reputation, spawned by the North Korean bomb test. I confess that I have not read all of the pieces he cites, but the first four(Machiavelli, Schilling, the stuff from the Cuban Missile Crisis, Mercer) are top-notch stuff. Mercer's work on Reputation and the fundamental attribution error really shaped my thinking--that we tend to explain the behavior of allies and enemies differently, reinforcing our pre-existing beliefs. Allies can do no wrong as we ascribe their bad behavior to circumstances and their desired behavior to their inherent goodness. Enemies, as Drezner summarizes, can do no right--if they do what we want, it is because of the context; and if they behave badly, it is because of their nature.
Efforts to establish reputation are, as Drezner's post suggests and Mercer convincingly argues, fraught with difficulties, so my simplistic recommendation would be--do only those actions you want to do for their essential qualities and expected effects, independent of their reputational consequences. Fight in Vietnam if it is in your interest, not because of the message it sends to anyone else. And not if it ain't.