[Json] Radically changing 4627bis

[Paul Hoffman <[e-mail]>]

<no hat>

On Oct 8, 2013, at 5:38 PM, "Manger, James H" <[e-mail]> wrote:

> But the publication of ECMA-404 means we should radically change 4627bis.
> Lets do what ECMA-404 "Introduction" suggests other standards do:
> 
>  It is expected that other standards will refer to this one, strictly
>  adhering to the JSON text format, while imposing restrictions on various
>  encoding details. Such standards may require specific behaviours. JSON 
>  itself specifies no behaviour. 
> 
> Lets reference ECMA-404 -- not repeating any of the syntax.
> Define the media type.
> Discuss the interop issues with surrogates, huge integers, decimal floats, duplicate names, comments, encodings.
> Perhaps define terms such as "web-safe JSON" (for a subset of JSON) and "web-safe JSON parser" (for specific parser behaviour) that other specs using JSON can refer to.
> 
> I guess this is close to Tim’s "Internet JSON" or "I-JSON" proposal.
> It is time to drop the minimal 4627 update.

A strong -1 on ripping out our syntax and pointing to ECMA-404. Our syntax matches their syntax, so there is not a conflict issue. ECMA-404 was supposed to be syntax-only, but they included semantics about Unicode (characters vs. code points) that were important enough to this WG to generate hundreds of messages on our mailing list. If we point to ECMA-404 for the syntax only, and add in our semantics on interoperability, there will then be a conflict that will have a negative effect on developers who are trying to create interoperable implementations.

--Paul Hoffman