-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 On 1/23/15 12:12 PM, Mike Jones wrote: > Periods are used in some JSON values, including strings and > floating point numbers, and so wouldn’t be a safe delimiter in all > potential future cases. What if your kty value was > “EC.little-endian”, for instance? > > >
I understand your points, but I still think it's more complex than it needs to be, and can lead to interoperability problems. Having an input that looks like you serialize a JSON object is going to lead some implementers to the false optimization of actually serializing a JSON object, regardless of what this document says. It might even work today for some JSON implementations -- and fail in a future update to that same implementation. I'd really rather see some other hash input generation scheme. - -- - - m&m Matt Miller < [email protected] > Cisco Systems, Inc. -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.22 (Darwin) Comment: GPGTools - https://gpgtools.org iQEcBAEBCgAGBQJUwqCNAAoJEDWi+S0W7cO1aAUH/2k8G1YW/8wpLegA9yj/Th5R 1Agwyks4X6nHQ9PLq+Mpv7cChrIAfDm1HRSKeJC7r/s5tUNfTpgeIwF9ChUQe2jt NbNp5Y5SUFdGxhx4qTwVrWiJT2k1ax83HMoy7D6aIXlmN5G/aBasR4/BCZKoPWEx MONiQd0ptPUUZkRCDtZtjjDN3AhFfbLJV4h6x7Z55S3Vh82GMgCfGNcqlqmIVB7M AZG4rEMczTOMOjgRnXKUUPxf3aaGCLfiG1zHY7YRbOtcjP6/1yq+wztNoHIyUUuF lHVTw2uAMtJLE1xI5u27tkvahNAhYuJHRnOYZh/a9AuwTjr3XbR56lmnSJXFwsQ= =oMOM -----END PGP SIGNATURE----- _______________________________________________ jose mailing list [email protected] https://www.ietf.org/mailman/listinfo/jose
