Improved documentation of exit-logging of states.

I got questions in email that made clear that the documentation
does not explain the meaning of the second logging trigger
(like the "/@" in WAIT(@/@)) well enough.
Hope it is better now.

org.el

@@ -1716,7 +1716,9 @@ that the WAIT state can be selected with the \"w\" key. \"WAIT(w!)\"
 indicates to record a time stamp each time this state is selected.
 \"WAIT(w@)\" says that the user should in addition be prompted for a
 note, and \"WAIT(w@/@)\" says that a note should be taken both when
-entering and when leaving this state.
+entering and when leaving this state.  The last double-setting is
+only a backup, to force a note even if the target state has no
+logging configured.
 
 For backward compatibility, this variable may also be just a list
 of keywords - in this case the interptetation (sequence or type) will be

org.texi

@@ -4,7 +4,7 @@
 @settitle Org Mode Manual
 
 @set VERSION 5.22a
-@set DATE February 2008
+@set DATE March 2008
 
 @dircategory Emacs
 @direntry
@@ -3015,9 +3015,12 @@ the @samp{Closing Note}.}, and that a note is recorded when switching to
 WAIT or CANCELED.  The setting for WAIT is even more special: The
 @samp{!} after the slash means that in addition to the note taken when
 entering the state, a time stamp should be recorded when @i{leaving} the
-WAIT state.  This allows you to get a record when switching from WAIT
-back to TODO, without getting one when first turning an entry into a
-TODO.
+WAIT state.  This second logging setting only comes into play when
+switching to a state that has no logging set up.  So it has no effect
+when switching from WAIT to DONE, because DONE is configured to record a
+timestamp only.  But when switching from WAIT back to TODO, the
+@samp{/!} in the WAIT setting now triggers a timestamp even though TODO
+has no logging configured.
 
 You can use the exact same syntax for setting logging preferences local
 to a buffer: