OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

search-ws message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [search-ws] Re: ATOM


Ray Denenberg, Library of Congress wrote:
> From: "Ashley Sanders" <a.sanders@manchester.ac.uk>
>   
>> What about creating a new operation called "feed" -- the only
>> difference to the current "search" operation is that the response
>> is in ATOM format. The "search" operation would be as it is now. I
>> can see objections to having two such similar operations,
>>     
>
> My first impression was exactly that: objections to two such similar
> operations.  But thinking more about it, maybe they wouldn't be quite so
> similar.   I think it's an idea worth some thought.
>
>   

The feed aspect is only a side effect of ATOM. What I find compelling 
about it is that is provides a ready made standard format for returning 
a search result and is extensible to handle any special needs of our spec.

Lets us see what we can agree upon...

I assume that we agree that a new ATOM representation would do 
everything the current representation does and provide new value on top 
of that:

    * supports feed
    * more RESTful
    * allows links to non-XML records
    * supports related links
    * supports metadata such as author, created, updated, title
    * supports extensiblity to handle custom needs of our spec

I assume that we also agree that not having a "required standard / 
default response representation" is unacceptable because it sacrifices 
interop.

At this point perhaps our view begin to diverge:

    * I favor that the "required standard / default response
      representation" be ATOM 1.0 and not the  current SRU response
      representation
    * Some others favor the current SRU response representation.

I have already shared my rationale for why ATOM should be the "required 
standard / default response representation".
I respectfully ask, what are the arguments in favor of the current SRU 
response format to be the "required standard / default response 
representation"?

<aside>
As for "operations" (new or old), reminder that REST is focused on 
"resources" and not "operations".
Another important REST is that of a "uniform interface" where the 
operations are fixes by the HTTP protocol (GET, PUT, POST, DELETE)
and these operations provide a consistent and uniform (fixed) interface 
that operates on any resource.

In REST terms we are talking about alternate representations of the same 
resource (results of a search algorithm).
<aside>

-- 
Regards,
Farrukh





[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]