Re: JSON-LD vs JWT for VC

> On Nov 3, 2018, at 21:20, Anders Rundgren <anders.rundgren.net@gmail.com <mailto:anders.rundgren.net@gmail.com>> wrote:
> 
>> If we also assume non malicious senders, there still remains a case for canonical data on the wire, that hopefully is compact, easy and fast to parse.
> 
> This is more problematic.  For JSON-LD-signatures (I'm not versed in JSON-LD so I may be off-track here) sending the canonicalized data may not be realistic since it is not JSON.
> 
> For Canonical JSON it is technically feasible but not necessarily not what you want:
> https://tools.ietf.org/html/draft-rundgren-json-canonicalization-scheme-01#appendix-C <https://tools.ietf.org/html/draft-rundgren-json-canonicalization-scheme-01#appendix-C>

I realize that “canonical” was poor wording on my part, especially in this thread. “Immutable" would have been a better word, but still somewhat incomplete. I’m neither referring to canonical as in JSON-LD nor in Canonical JSON.

There was a debate in CBOR-land about how canonicalization should fit into the data description language and as far as I know, it is most likely not going into the data description language, but there seems to be interest to define canonicalization methods. There is probably not going to be one method to fit all purposes.

For example, network packets better have the destination address in the front and integrity in the back.
The DID spec, like any sane spec presents types before values in the examples.
Lexicographical ordering might break reasonable ordering for processing for both of those cases.