[occi-wg] Today's confcall, items to the agenda

Gary Mazz garymazzaferro at gmail.com
Tue Oct 2 10:10:38 EDT 2012


Great job...

Comments in line

On 10/2/2012 3:57 AM, Ralf Nyren wrote:
> Hi,
>
> I would like to add the following items to the agenda.
>
> Split of Rendering into Protocol & Data Format
> ----------------------------------------------
> Where to describe this separation in the documents?
This should go into the rendering document...
>
> OCCI Core?
> The preamble (include/introduction.tex) included in all documents?
> ...
Agreed. These documents require a description of the specifications and 
the role the specifications in the specification hierarchy.
>
> Pagination
> ----------
> Is the JSON rendering ready for pagination?
>
> If we follow the OpenStack approach to pagination (in the future version
> of the HTTP rendering doc) almost everything can go into the HTTP request
> parameters. All good.
>
> However, the ability for a client to distinguish between a full collection
> response and a partial paginated response is IMO still unresolved.
>
> Adding a "total collection size" item in the JSON response is one
> solution. Another solution is for the client to ask an extra request to see
> if there is more data to retrieve.
I don't like the approach taken with openstack for numerous reasons. 
There is little understanding of query subtleties and the greater issue 
of query interface and capability.

>
> Action invocation
> -----------------
> Since the JSON payload is designed to be usable with multiple transport
> protocols I think it is important that you always can determine what a
> certain JSON payload corresponds to.
>
> How do you distinguish between an Action invocation and a discovery
> interface response?
>
> If the payload contains an "action" object it is an invocation. If the
> payload contains an "actions" (plural) it is a discovery response.
>
> So, IMO it works as it is but the difference is indeed subtle.
This goes back to my proposal/discussion to have OCCI data (JSON) as a 
standalone document, ala OVF.  As a stand alone document, occi object 
life cycle ACTIONs are included in the document. The exact structure and 
capabilities of a standalone document needs to be defined. A standalone 
OCCI document can support a more robust semantic for ACTIONs simplifying 
OCCI administration tasks.   A standalone OCCI document can also bring 
OCCI ACTIONS to less robust protocols, ie. messaging.
> Attribute properties
> --------------------
> Have you all looked at the Attribute type in OCCI Core errata? Is
> everything there now?
>
> I remember Florian talked about an optional "description" property. Should
> we add that as well?
What is the verdict on the core document, a new comment phase or errata ?
>
>
> regards, Ralf
>
>
> On Tue, 2 Oct 2012 08:59:46 +0000, "Metsch, ThijsX"
> <thijsx.metsch at intel.com> wrote:
>> Hi all,
>>
>>
>>
>> Important call today on JSON and the OCCI core document! Please join at
>> 4pm UTC, 5pm GMT, 6pm CET.
>>
>>
>>
>> Numbers:
>>
>> Austria 0820 4000 1503
>>
>> Belgium 070 35 99 45
>>
>> Canada 1 213 289 3444
>>
>> France 0821 230 748
>>
>> Germany 01803 001 178
>>
>> Ireland 0818 270 007
>>
>> Italy 848 390 166
>>
>> Netherlands 0870 001 909
>>
>> Poland 0801 003 533
>>
>> South Africa 087 550 0375
>>
>> Spain 902 885 318
>>
>> Sweden 0939 2066 400
>>
>> Switzerland 0848 560 190
>>
>> United Kingdom 0844 4 73 73 73
>>
>> United States 1 415 363 0833
>>
>>
>>
>> More: http://pdf.powwownow.com/pdf/USA_en_pwn-dial-in-numbers.pdf
>>
>>
>>
>> PIN code: 698133
>>
>> -------------------------------------------------------------
>> 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
> https://www.ogf.org/mailman/listinfo/occi-wg



More information about the occi-wg mailing list