MIME Type Review Request: text/CSTA-type

[Chris Lilley]

On Saturday, November 6, 2004, 6:22:11 PM, Bruce wrote:

BL> On Tue November 2 2004 08:54, Hollenbeck, Scott wrote:
>> Please review the MIME type registration template described below.  The IESG
>> has received a request to register this MIME type in the standards tree.  It
>> is a product of Ecma International.  A URL for the formal specification is
>> included in the template.
>> 
>> -Scott-
>> ----------
>> MIME media type name: text
>> 
>> MIME subtype name: CSTA-type
BL> [...]
>> Security considerations:
>> This content type is designed to carry CSTA data types over network
>> protocols. Appropriate precautions should be taken to insure that
>> applications observing these CSTA objects are authorized to do so.
BL> [...]
>> Applications which use this media type:
>> The text/CSTA-type MIME type is used to carry CSTA data types specified in
>> CSTA XML (ECMA-323) over various types of network protocols.
>> 
>> Additional information:
>> CSTA XML (ECMA-323) is an application level protocol that enables an
>> application to control and observe communications involving various types of
>> media (voice calls, video calls, instant messages, Email, SMS, Page, etc.)
>> and devices associated with the media.

BL> Why is this being proposed for registration in the text media type tree?
BL> As far as I can tell, this media type appears to be an application data
BL> format containing control data, not textual information (in the sense
BL> of the use of "text" in the MIME standards (RFC 2046) and Internet
BL> Best Current Practice (RFC 2277):

In addition, its an XML media type so

- it should use the +xml  convention
- it should not use text/*

The latter is currently allowed by RFC 3023 but that is being revised to
deperecate use of text/* for XML types.

So, more reasons it should be application/CSTA-type+xml


BL> I can see nothing in the proposed media type registration or in the
BL> referenced document that indicates that the media type is to be
BL> used to convey human-readable natural language text as
BL> opposed to application-specific data type,

Agreed

BL> nor do I see any provision
BL> for carrying information about language of text as required by
BL> RFC 2277.

(If its xml, that would be done with xml:lang attributes)

BL> A concrete example of how the proposed media type would be
BL> used to convey human-readable text in some specified language
BL> would help to provide justification for registration of the
BL> proposed subtype in the text media type tree.
 
And even then, I would argue it should not be in the text tree.



-- 
 Chris Lilley                    mailto:chris at w3.org
 Chair, W3C SVG Working Group
 Member, W3C Technical Architecture Group