Julian Mehnle wrote:
I thought it up myself
Hm, not good enough, they are very paranoid with RfC 2822, I
need something like Postel, Knuth, Dijkstra, or Benedict XVI.
I'm sure you can find that rule in any good software
engineering textbook.
My books are old, when I learned this stuff the generation of
test suites from formal spec.s and indeterministict abstract
data types were still _research_ (= not in just any good book).
I hope they didn't torture you with the old "stack" examples.
OB on topic: Don't add a plain "result" to the ballot, it
would have too many side effects in other parts of the spec.
Bye, Frank