Re: PaceContentNegotiationSection

I'm sorry you feel that way, I'm just an implementor who is having difficulty 
with your spec and is trying to decipher your wording, particularly this:

"The  content  of  this  document  can  vary 
        based  on  aspects  of  the  client  request,  including,  but  not 
        limited  to,  authentication  credentials. "

Why not, "the document returned may vary" instead of confusing me by stating 
that the _contents_ of the document may vary, instead of the _location_ of 
the document may vary.  Seems to indicate the preferred implementation is to 
have /service respond differently to each user as you've shown in your 
example, not redirect, or send a different service document for different 
users... no, it quite clearly states there is to be one document whose 
content, not location, is to vary by user...

Anyone else see where this just might be a tad confusing?

-EJB

>-----Original Message-----
>From: James M Snell [mailto:jasnell-Re5JQEeQqe8AvxtiuMwx3w@xxxxxxxxxxxxxxxx]
>Sent: Friday, August 18, 2006 07:47 PM
>To: eric-MkmoNbc1SAncr/OS1auqaA@xxxxxxxxxxxxxxxx
>Cc: atom-protocol-O6w3ZxSwtmQ@xxxxxxxxxxxxxxxx
>Subject: Re: PaceContentNegotiationSection
>
>Sigh... this is going nowhere.  The spec defines a service document
>format ands says that service docs can be retrieved by GET on some URI,
>that's it. That's all it says. Nothing more.  Everything else is your
>decision. How you model your resources is your decision. Whether or not
>to use conneg is your decision. Whether to serve one service doc or 50
>from one uri or 50 uris is your decision.  The spec is not broken just
>because the choices you've made on how to implement your server make it
>hard for you to do so.
>
>I'm done with this conversation now.
>
>- James
>
>Eric J. Bowman wrote:
>>> Great, we haven't spec'd that all introspection docs you serve from your
>>> server MUST have the same URI.  What URI(s) you choose to server your
>>> introspection docs from is your own concern.
>>>
>> 
>> No, what you've specced in is that all introspection URIs may serve multiple 
>> resources which happen to all be service documents.  This is a bad thing 
>> because these documents are not each representations of the same resource, 
>> they are each representations of unique resources.
>> 
>> Back to my original objection about the spec wording, "THE document".
>> 
>> -EJB
>> 
>> 
>> 
>> 
>