Re: [Json] I-JSON Tpic #2: Top-Level

On Mon, Apr 28, 2014 at 3:50 PM, Matt Miller <mamille2@cisco.com> wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA512
>
> "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."
>
> If you are in favor -- or not -- of restricting the top-level I-JSON
> text to only be object (or array / object), please respond to this
> thread with your reasoning.
>

I am not in favor of restricting the top-level I-JSON text only to be an
object. This would make it stricter than RFC4627/ecma-262 and then stand to
muddy the waters of what the application/json media type currently states.
I don't believe it is a good idea to make this stricter, without a matching
media type, which I think makes this also a non-starter.

I am content to see it as array or object at the top level. Since this
seems to be the current level of accepted use by most people within the
JSON community due to RFC4627.

But, I would greatly prefer that any JSON value be allowed as the top level
item. The requirement that any JSON value be represented within a structure
creates an amount of unneeded indirection and overhead.

If my program intends to set a local value to be the response from another
source. It should not have to know that it must descend into the actual
result to obtain the sources intended value. If both sides agree to send
array/object structures, akin to golang's multiple return values, I
consider this a better practice. But it should not be a requirement, or an
imposed overhead.

Add to this, http://www.ecma-international.org/ecma-262/5.1/#sec-15.12.1.2seems
to define JSON text as any JSON value in contrast with RFC4627. My
reading/understanding of RFC4267/ecma-262 may be incorrect, but my feeling
that restrictions on the top level value should be lifted, not tightened.

>
>
> - --
> - - m&m
>
> Matt Miller < mamille2@cisco.com >
> Cisco Systems, Inc.
>
> * "How to Argue about I-JSON" <
> http://www.ietf.org/mail-archive/web/json/current/msg02775.html >
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG/MacGPG2 v2.0.22 (Darwin)
> Comment: GPGTools - https://gpgtools.org
> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
>
> iQEcBAEBCgAGBQJTXrEZAAoJEDWi+S0W7cO1bjAH/RxLeWEABXMoJEJrw92lYfkP
> bLgoS7TRoSxVdUY9K8SVMEJKHMrY2QrK/aNqhwlpWvIirM70xFUT7Sb40M+Msmm/
> qFZTFfVRdRsnqsd21dciZ6hA4eSDfkP81cIoMn4MiCx8EExZCj7yroWaTor8jNct
> 7BgVCWkom2ENVNOb6JEPFa7LD6hpm/9vx1FHctbDgreJuhsYxY9Uy17h/BjxxwPc
> NkNFbo+OX4Zw6f053cKWbITps/R+IAM57atzJ9PgGpMpN+POQJF/BihnB1qRdUsv
> y8QqQ1s8r47DauWQ0eX4BxgPXeGV+jBqXrtVAzyNq2B7KFkDukf9UVahUaJjwJU=
> =nAfX
> -----END PGP SIGNATURE-----
>
> _______________________________________________
> json mailing list
> json@ietf.org
> https://www.ietf.org/mailman/listinfo/json
>

-- 
Matthew P. C. Morley