Re: [ietf-types] Q's regarding development of +json media types

Nathan <> Thu, 30 September 2010 21:03 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 1A6873A6E77 for <>; Thu, 30 Sep 2010 14:03:33 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -1.855
X-Spam-Status: No, score=-1.855 tagged_above=-999 required=5 tests=[AWL=0.745, BAYES_00=-2.599]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id LpsaJdaKKOGi for <>; Thu, 30 Sep 2010 14:03:31 -0700 (PDT)
Received: from ( []) by (Postfix) with SMTP id B3ECA3A6E37 for <>; Thu, 30 Sep 2010 14:03:31 -0700 (PDT)
Received: (qmail 3619 invoked from network); 30 Sep 2010 21:04:17 -0000
Received: from unknown ( by ( with ESMTP; 30 Sep 2010 21:04:15 -0000
Message-ID: <>
Date: Thu, 30 Sep 2010 22:03:42 +0100
From: Nathan <>
Organization: webr3
User-Agent: Thunderbird (Windows/20100228)
MIME-Version: 1.0
To: Gregg Kellogg <>
References: <> <>
In-Reply-To: <>
Content-Type: text/plain; charset="ISO-8859-1"; format="flowed"
Content-Transfer-Encoding: 7bit
Cc: "" <>
Subject: Re: [ietf-types] Q's regarding development of +json media types
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: "Media \(MIME\) type review" <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 30 Sep 2010 21:03:33 -0000

Hi Gregg,

Indeed one is a simple json based serialization of RDF which is very 
similar to N-Triples, and thus not specifically a serialization of 
N-Triples as you say but rather a simple JSON based serialization of RDF 
that shares many similarities with N-Triples, the same feature set if 
you will.

The N3-like JSON serialization (JSN3) started out life as 
yet-another-json-rdf serialization which was Turtle-like, however soon 
had Graph Literals added at the request of Tim Berners-Lee, who then 
swiftly followed up with another request to make it fully round-tripable 
with Notation-3 and thus to express/serialize everything Notation-3 
does. Similarly this now means that a subset of it could be used as a 
Turtle-like JSON serialization of RDF, or even an AMORD-in-RDF-like 
serialization (which is basically Turtle + Quoting) - Will cross that 
bridge later though!

With regards relating to JSON-LD, in some ways yes, as in Manu and I 
have discussed at length and gone over the design choices of both 
JSON-LD and JSN3, and also the various other JSON based serializations 
of RDF which are in the wild & feedback has been gained from many who 
work heavily with RDF, N3 and JSON - and of course much more expected 
and welcomed!



Gregg Kellogg wrote:
> I don't think it makes sense to describe JSON serializations of Notation 3 or N-Triples, as these are themselves just serializations of RDF. Perhaps what you're describing is two different JSON serializations of RDF, one which is more closely related to Notation-3 and the other N-Triples. Although, note that Notation-3 is, itself, a super-set of N-Triples (as is Turtle [1]), so is there really a requirement for two different serializations, or isn't it up to the author to chose a particular syntax using different capabilities of the format.
> Also, note that Notation-3 is more than a syntax, as it includes behavior as well as representation, which is one reason Turtle was created. A purely syntactic version of Notation-3 is n3-rdf [2], which is a relatively small super-set of Turtle itself.
> Does this relate, in anyway to JSON-LD [3], which is also an RDF serialization in JSON?
> Perhaps application/rdf+json would be an appropriate mime-type.
> Gregg
> [1]
> [2] See N3-rdf at
> [3]
> On Sep 30, 2010, at 12:27 PM, Nathan wrote:
>> Hi,
>> I'm currently working on / experimenting with two JSON based 
>> media-types, the first is a JSON based serialization of Notation 3, the 
>> other a JSON based serialization of N-Triples.
>> I've got to the point now where I have implementations of both and need 
>> to content-negotiate over them, thus my initial question is what media 
>> type should I use in the interim whilst experimenting?
>> For a later date, if I were to move towards seeking registration, would 
>> the best approach be to work on the specifications out with any 
>> standards body, or to do the work as an internet-draft, or?
>> Would the recommended approach be to work towards a json-schema approach 
>> or towards a "+json" type?
>> And finally, for JSON based media-types do you have any special 
>> considerations or gotcha's I should be taking in to account at an early 
>> stage?
>> Best & TIA for any response,
>> Nathan
>> _______________________________________________
>> ietf-types mailing list