On Wed, 1 Jul 2009, Ralf Baechle wrote: > Code written to require intimate knowledge of all operator priorities is > probably as bad as overuse. Here a few examples: > > return (1); > #define FOO (42) > a = (b + c); > a = (b + c) + d; > a = (b * c) + (d * e); > if ((a > b) && (a < c)) Sometimes additional brackets are useful to emphasise some logical grouping of entities used, so I would accept even such a construct as: a = (b + c + d) + (e + f); to bring attention of the reader that a, b and c refer to something other than e and f, but yes, it would require explicit justification and I wouldn't consider it a blanket accept statement for such kinds of expressions. The #define case makes my eyes hurt. OTOH we require unnecessary brackets in the context of sizeof. ;) Maciej
