values of Type

[Drummond Reed]

Sorry for the delay -- was out at my son's soccer tournament tonight.

Answers:

1) In the XRD schema, the Type element takes a value of type "anyURI".
However it is important to note that according to the W3C XML spec, this can
be either absolute or relative. So *technically* Andy's example is not
incorrect.

However I fully agree that Type values SHOULD be absolute in order to reduce
the potential for collision. So Andy's examples SHOULD have
"+authentication" and "+photos", as these are absolute XRIs. (Note that
these are equivalent to "xri://+authentication" and "xri://+photos", i.e.,
the "xri://" part is optional with absolute XRIs that start with global
context symbols (=, @, +, $, !. This was conscious design decision to
simplify human usability, e.g., to type a global XRI in an editor or address
bar only requires a single symbol character.) 


2) The purpose of the MediaType element does not overlap with the purpose of
the Type element (and Andy's example of "isso" as a media type is incorrect
-- we're talking IANA-registered media types here, like "text",
"application/xml", etc.). In XRI resolution (and potentially for other
applications using XRDs), MediaType is just further metadata available to
describe the capabilities of a service endpoint.

Hope that helps.

=Drummond 

-----Original Message-----
From: Johannes Ernst [mailto:jernst+lists.danga.com at netmesh.us] 
Sent: Tuesday, December 06, 2005 2:06 PM
To: Drummond Reed
Cc: Dale Andy; YADIS List
Subject: values of Type

Drummond,

I'm just looking through Andy's XDI slides from his presentation  
yesterday that I unfortunately had to miss. A question arises.

One of his examples contains the following XML fragment.

<service>
	<type>authentication</type>
	<URL>http://www.idp.com/</URL>
	<mediaType>isso<mediaType> 	
<service>
<service>
	<type>photos</type>
	<URL>http://www.kodakgallery.com/AlbumMenu.jsp</URL>	
	<mediaType>img<mediaType>
<service>

Which, apparently and not surprisingly, instantiates the same XML  
format as we do in YADIS. However, it appears to me that there are  
two significacnt distinctions in use:
  - YADIS assumes that the value of "type" is a URL, such as http:// 
lid.netmesh.org/sso/1.0
  - Andy's slides seem to assume that "mediaType" (not used in YADIS)  
is a very substantial attribute that, in this use, seems to partially  
overlap with what we use the URL for for YADIS purposes.

So ... this poses some or all of the following questions:
1) did I get it wrong?
2) does the current draft standards document prescribe one or the  
other, and if so which?
3) if it doesn't prescribe one or the other, does the document need  
to be tightened so interoperabilty is realistic?


Johannes Ernst