Re: [JSON] I say again, what *is* JSON?

On Thu, Mar 24, 2011 at 1:50 PM, Peter Frederick Patel-Schneider <
pfps@research.bell-labs.com> wrote:

> From: Nathan <nathan@webr3.org>
> Subject: Re: [JSON] I say again, what *is* JSON?
> Date: Thu, 24 Mar 2011 12:40:08 -0500
>
> [...]
>
> >> Not very well.  To understand JSON this way is extraordinarily difficult
> >> and expensive, requiring deep knowledge of the innards of EMCAScript.
> >
> > I had hoped it would a little, since it was in relation to multiple
> > "keys" with the same name and also confusion over that terminology in
> > relation to JSON, however, I guess it would be good to know how you see
> > this negatively impacting the JSON TF and JSON related work of the
> > group? Or specifically what are your current issues and how do they
> > affect the work we're planning/doing?
> >
> > Best,
> >
> > Nathan
>
> Well, I, for one, find it hard to work on standardizing against anything
> when I don't know the target.   And not just the general ideas of the
> target, but also all the corner cases, etc., etc.
>
> To pick a particular example, I was happy to see that http://json.org/
> (which I thought would have been, if not definitive, at least quite
> close to correct) allow for multiple values for the same pair name in JSON.
> I was also not expecting to have the EMCAScript document be confused as
> to how to handle multiple values for the same pair.
>
> To pick another example, I wasn't expecting to see the relationship
> between objects in JSON and objects in the host language to be so subject
> to extra inflences.  (Not that the possibility of these extra influences
> are bad per se, but they sure affect how I view JSON.)
>
>
> So, I'm still confused as to exactly what JSON is.
>

Respectfully speaking, I don't buy your assertion that the WG can't proceed
without a crystal-clear understanding of all the intricacies or gaps in the
various specs.

The fact that there's confusion or ambiguity among some of the specs around
JSON and Javascript hasn't stopped it from becoming a very useful tool for
the web development community, and I don't think think it should stop us
from coming up with something useful.

If it helps, think of JSON as a collection of best practices for serializing
simple key/value data structures from a variety of target languages into a
Javascript syntax.  That's still a very useful thing to have even if it
isn't "standardized".  As long as you stick with simple datatypes and data
structures, you'll find pretty universal tool support for what you need to
do.  If you're worried about standard behavior on some of the corner cases,
then stay away from those corner cases.

It shouldn't be the task of this WG to sort out all of the specs to come up
with a "standard" JSON where there isn't one already.  IMHO our task is to
identify and promote some best practices for using JSON with RDF; if the
lack of a clear and unambiguous standard bothers you then maybe the WG can
produce a working note as opposed to a recommendation.

-Alex