>>> parsers need to canonicalize maps to any depth in order to >>> detect duplicates. This is "complex" by any definition of the word. It isn't complex in terms of computational efficiency ... you can canonicalize in O(N log N) and do it while reading. And the consequence of not using structure-equality in duplicate detection is complex. > I think CBOR should be clear about how it handles sharing and equality. agree. Larry -- http://larry.masinter.net (obscure footnote: Serializing structures with duplicate pointers is fun, you need some notation to signify back pointers and to drop anchors. Using hash tables and canonical values was part of the circlprint/hprint http://pdp-10.trailing-edge.com/decuslib20-01/01/decus/20-0004/21lisp.tty.html implementation.)