[occi-wg] OCCI Categories and Types

[Gary Mazz]

Hi,

I'm not concerned about using "rel" per se, it the relationships 
considered permissible by OCCI which is the concern. Additionally, we 
need to express the elements of the we intend to use for OCCI. 
Application of method through examples is a good first step, but 
requires further definition in the specifications. Simply referencing 
RFCs may not be adequate to ensure and sustain interoperability.  

Referencing the section 6.2.2 in Rfc [1], it defines a registry  of 
relationships for the 'rel'  attribute.  Outside of the ordered 
sequences, we seem to be defining new relationships  or at least 
relationships undefined by the rfc [1]. The out of specification [1] 
relationships we need to define in the OCCI specification.

I think we need to have a discussion on occi extensibility;  method and 
form.

BTW, I'm running into the similar issues with the console proposal. I 
have both credentials and security identifiers (SIDs) that fall into the 
extensibility issue.

cheers,
gary


Edmonds, AndrewX wrote:
> RE: 2)
>
> It's not repurposed by OCCI rather the web linking specification [1]. The attribute "rel" is defined by the web linking specification. See page 6 of [1]. 
>
> HTH,
>
> Andy
>
> [1] http://tools.ietf.org/html/draft-nottingham-http-link-header-10
>
> -----Original Message-----
> From: Gary Mazz [mailto:garymazzaferro at gmail.com] 
> Sent: Monday, August 16, 2010 6:24 AM
> To: Edmonds, AndrewX
> Cc: occi-wg at ogf.org
> Subject: Re: [occi-wg] OCCI Categories and Types
>
> Hi,
>
> Great write and example..
>
> I have a couple of comments not covered by Ralf Nyren's emails.
>
> 1) The use of multiple schemes and multiple attributes, permissible in 
> headers may cause significant rendering and referential issues in the 
> body.  I'm fairly confident most parsers will have problems with this 
> approach.  If this scheme is also intended for a body, I'll create and 
> example to examine the doms
>
> 2) The role of the "rel" attribute is undefined in the specification. In 
> one example it is used as a Link attribute indicating the Link's 
> Resource relative position in a sequence. In the GET response example, 
> 'rel' is repurposed to extend the catagory of the Link. Rel will also be 
> used in the RDFa representation possibly with a role differing from the 
> examples. This will ultimately add to confusion for both the HTTP and 
> RDFa implementations.
>
> I think before we start producing examples using extensions, we should 
> define the types of extension permitted and define where they are placed 
> in the model scheme.
>
> -gary
>  
>
> Edmonds, AndrewX wrote:
>   
>> Hi all,
>> Yesterday Alexander and I spent some time exploring and discussing OCCI categories. These categories make up the type system for OCCI and I've documented our discussion here [1]. We'd love to hear feedback, comments etc!
>>
>> Andy
>>
>> [1] http://forge.gridforum.org/sf/wiki/do/viewPage/projects.occi-wg/wiki/Category
>>
>> -------------------------------------------------------------
>> Intel Ireland Limited (Branch)
>> Collinstown Industrial Park, Leixlip, County Kildare, Ireland
>> Registered Number: E902934
>>
>> This e-mail and any attachments may contain confidential material for
>> the sole use of the intended recipient(s). Any review or distribution
>> by others is strictly prohibited. If you are not the intended
>> recipient, please contact the sender and delete all copies.
>>
>> _______________________________________________
>> occi-wg mailing list
>> occi-wg at ogf.org
>> http://www.ogf.org/mailman/listinfo/occi-wg
>>
>>   
>>     
>
> -------------------------------------------------------------
> Intel Ireland Limited (Branch)
> Collinstown Industrial Park, Leixlip, County Kildare, Ireland
> Registered Number: E902934
>
> This e-mail and any attachments may contain confidential material for
> the sole use of the intended recipient(s). Any review or distribution
> by others is strictly prohibited. If you are not the intended
> recipient, please contact the sender and delete all copies.
>
>
>