Re: Last call comment: Why?

On Thu, 18 Nov 2004, Yalcinalp, Umit wrote:
> 
> The usage of contentType attribute is indeed optional. It is not 
> required per section 3. [1]. Therefore, we are not clear why you think 
> there is a problem.

Well I wasn't really saying there was a problem, per se, I'm just curious 
as to what the expected use cases are for this attribute.

For example, in the Web Forms 2 spec that I am working on there is a part 
that defines an XML document language that specifies additional media type 
information for binary data [2]. According to the "Assigning Media Types 
to Binary Data in XML" spec, the WF2 spec SHOULD denote this by using a 
an element information item defined with an optional mime:contentType 
attribute and the content defined as base64-encoded data (or hex binary).

But I don't understand why WF2 should use that definition, instead of 
using just a local "type" attribute and only allonwing base64 content.

I guess my suggestion would be to change the SHOULD in section 3 to a MAY.

However, I am still curious as to why anyone would ever want to use this 
namespaced attribute instead of just using a local attribute. What are the 
expected advantages of doing so?

[1] http://www.w3.org/TR/xml-media-types/#usage
[2] http://whatwg.org/specs/web-forms/current-work/#x-www-form-xml

Cheers,
-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Friday, 19 November 2004 15:51:48 UTC