This is the first time I've seen Preserves and I appreciate that! I've written something much simpler before [1] and understand that the specification writing is a huge undertaking. Please allow me to ask some quick questions (which answer I may have read but missed):
- Given the popularity of JSON among others, you will be almost certainly in need of the explicit JSON encoding of Preserves data model. Probably also others.
- Why a separate binary syntax? Unlike the textual format and semantics I don't see much value in yet another binary serialization format, given CBOR seems moderately successful and does have a well-defined data model. It seems that CBOR plus some already registered tags suit the needs for binary Preserves.
- It is fragile to base the textual syntax on an unspecified Unicode version. ECMAScript specifies the minimum required version (ES5 requires 3.0.0, ES2015 requires 5.1 and ES2016 requires 8.0) and XML (and also CSON) specifies identifier syntax in the way that the future version of Unicode doesn't invalidate that. I believe the latter is more suited for serialization uses.
This is the first time I've seen Preserves and I appreciate that! I've written something much simpler before [1] and understand that the specification writing is a huge undertaking. Please allow me to ask some quick questions (which answer I may have read but missed):
- Given the popularity of JSON among others, you will be almost certainly in need of the explicit JSON encoding of Preserves data model. Probably also others.
- Why a separate binary syntax? Unlike the textual format and semantics I don't see much value in yet another binary serialization format, given CBOR seems moderately successful and does have a well-defined data model. It seems that CBOR plus some already registered tags suit the needs for binary Preserves.
- It is fragile to base the textual syntax on an unspecified Unicode version. ECMAScript specifies the minimum required version (ES5 requires 3.0.0, ES2015 requires 5.1 and ES2016 requires 8.0) and XML (and also CSON) specifies identifier syntax in the way that the future version of Unicode doesn't invalidate that. I believe the latter is more suited for serialization uses.
[1] https://github.com/lifthrasiir/cson