Re: [apps-discuss] JSON Schema considered harmful

[Francis Galiegue <[e-mail]>]

On Thu, Sep 20, 2012 at 10:24 PM, Dave Crocker <[e-mail]> wrote:
[...]
>
> I was suggesting a small-but-useful subset, which I thought was nicely
> exemplified by the terse list of practical direction of the precipitating
> note about choosing a notation language.
>
> Very narrow topic.  Very important.  Often very religious and very confused.
>
> Just the stuff of a BCP, IMO...
>

I-D 3 of JSON Schema has it: a "small-but-useful subset", which can be
used to describe _any_ JSON value. It has sufficient intrinsic value
that it is being used, for instance, by the Google discovery API.

However, there was room for improvement. In particular, the fact that
the core schema keywords were not fully defined, and also section 5.27
which has been a constant source of misinterpretation. And the fact
that some "untold", but important stuff, was lurking behind. And this
is precisely the goal that this ad-hoc organization "of no value" is
pursuing, and why it takes time. But XSD is at best a goal NOT to
pursue.

And yes, all keywords are meant to define the _structure_ of JSON data
anyway. Semantic validation keywords exist, but the goal is to make
them optional.

In any case, I-Ds will be submitted via the official mechanism. Some
refining is needed before that.

-- 
Francis Galiegue, [e-mail]
JSON Schema: https://github.com/json-schema
"It seems obvious [...] that at least some 'business intelligence'
tools invest so much intelligence on the business side that they have
nothing left for generating SQL queries" (Stéphane Faroult, in "The
Art of SQL", ISBN 0-596-00894-5)