8 Some things notable, but not (yet?) covered by the manual.
11 - constants of result type prop should always supply concrete syntax
12 (elaborate on this in last sect of 'Defining Logics' (?));
14 - 'Variable --> Constant' possible during rewriting;
16 - 'trivial definitions' via macros (e.g. "x ~= y" == "~ (x = y)");
18 - patterns matching any remaining arguments are not possible (i.e. what would
19 be (f x y . zs) in LISP); e.g. HOL's @ (supposing it implemented via macros
20 which it is *not*): "@x. P" == "Eps(%x. P)", now the print rule doesn't
21 match things like Eps(%x. P, a, b, c);
23 - alpha: document the precise manner in which bounds are renamed for
26 - parsing: applications like f(x)(y)(z) are not parse-ast-translated into
27 (f x y z); this may cause some problems, when the notation "f x y z" for
28 applications will be introduced;