Re: [Json] How to argue about I-JSON

[Nico Williams <[e-mail]>]

On Mon, Apr 28, 2014 at 1:16 PM, Tim Bray <[e-mail]> wrote:
> Argument 2 - Top-level
>
> draft-i-json-01 says the top level is an object (but doesn’t say MUST, eek).
> I have heard arguments for allowing arrays too; anyone who thinks that is a
> good idea should speak up. (I won’t.)

This is really for indicating schema information in-band.  But we
already rely on indicating format out of band (via MIME type).  I find
in-band schema signalling in JSON to be a bad idea, especially when we
can do it out-of-band anyways.

Question: can schema information be encoded as a parameter to the
application/json media type?

> Argument 4 - Software behavior
>
> Section 3 of draft-json-01 implies draconian error handling - when a message
> is specified to be i-json but the receiver finds, for example, a dupe key,
> it is required to halt and catch fire. There’s a chance that this will be
> interpreted as “tbray poisons JSON with XML draconianism”. People with
> alternate language should suggest it. (I probably won’t argue this one
> either way.)

A bigger problem is that this requirement involves changes to existing
parsers to meet it.

Nico
--