Re: [apps-discuss] +exi

Carine Bournez <> Sat, 17 December 2011 10:41 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id A913C21F8C04 for <>; Sat, 17 Dec 2011 02:41:12 -0800 (PST)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -10.133
X-Spam-Status: No, score=-10.133 tagged_above=-999 required=5 tests=[AWL=-0.134, BAYES_00=-2.599, J_CHICKENPOX_33=0.6, RCVD_IN_DNSWL_HI=-8]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id KqPHk3aK3VeI for <>; Sat, 17 Dec 2011 02:41:12 -0800 (PST)
Received: from ( []) by (Postfix) with ESMTP id 34DC221F8BFE for <>; Sat, 17 Dec 2011 02:41:11 -0800 (PST)
Received: from carine by with local (Exim 4.69) (envelope-from <>) id 1RbrhP-0005ZE-G4; Sat, 17 Dec 2011 05:41:03 -0500
Date: Sat, 17 Dec 2011 05:41:03 -0500
From: Carine Bournez <>
To: Zach Shelby <>
Message-ID: <>
References: <> <> <> <> <> <> <> <> <> <>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <>
User-Agent: Mutt/1.5.18 (2008-05-17)
Cc:, Mark Nottingham <>, Thomas Herbst <>, "" <>
Subject: Re: [apps-discuss] +exi
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: General discussion of application-layer protocols <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Sat, 17 Dec 2011 10:41:12 -0000

On Fri, Dec 16, 2011 at 09:24:05AM +0200, Zach Shelby wrote:
> Right. It is exactly these kinds of applications that are using schema informed mode natively (without intermediate XML), where the application/foo+exi media type helps to determine this information out of band. When registering foo+exi, the schema information should be included in the registration.  In addition to SE2 mentioned so far, also our specification on SenML [] is requesting an application/senml+exi media type. Both of these media types are aimed at constrained embedded devices.

It seems that you could use application/foo and the content-encoding: exi
instead of registering a different media type. Is your application using a
protocol that has no content-encoding? or do you mean that you use a 
different schema for the exchange of a foo serialized in EXI than a 
text foo? A quick look at 
makes me think it is the latter, for SenML. It does not seem a very nice
way to convey the schema information, SchemaId is meant to serve this 
purpose, not the media type. Since you need to carry other EXI options,
out-of-band or in the EXI header, why not using SchemaId?.

Carine Bournez -+- W3C Europe