MS> Sure, but there's no guarantee that a given macro has hygienic MS> behavior. Generally, maintaining hygiene this way is pretty MS> tedious, which is why programmers usually don't go to the MS> necessary lengths. I will make a risky statement: a powerful tool should be used carefully. What you say is certainly true, but - as Paul Graham makes clear - macros are a way to fool with the way lisp evaluates your statements. To me, this is like a big red sign yelling: "Careful". I believe that even in the Elisp programming style tips people are advised not to use macros if they can be avoided. One example that has bitten me much to often was the useless use of macros in mailcrypt or in some VM file, IIRC, which broke whenever I installed a new VM version. Therewas absolutely no reason to use macros in that places. Your example below is of the same class: MS> (defvar bar 23) MS> and I want to refer to bar (that is, the above binding) in a MS> macro body: MS> (defmacro foo () 'bar) MS> ... but I do: MS> (let ((bar 42)) (foo)) => 42 Excuse me, but this is evil by design. Brr, what awful code. Holger -- --- http://www.coling.uni-freiburg.de/~schauer --- "Was für'n Agent??? Und wieso benutzt ihr nicht einfach alle die Standardlösung Outlook Express (früher MS News)? Ist IMHO das beste und übersichtlichste Programm für PC und dazu noch kostenlos!" -- Aus "Wie starte ich einen Endlosthread ?", Teil 1239