[foaf-protocols] First WebID Teleconference minutes (July 27th 2010)

Nathan nathan at webr3.org
Mon Aug 2 19:19:35 CEST 2010


Henry Story wrote:
> On 2 Aug 2010, at 19:09, Nathan wrote:
> 
>> Henry Story wrote:
>>> On 2 Aug 2010, at 18:45, Nathan wrote:
>>>
>>>> I really like that sentence - perfect even imho.
>>>>
>>>> 'MUST have a Machine and Human readable representation of a structured 
>>>> profile document'
>>> "MUST have a representation of an RDF graph in a machine readable representation. See the section on representations for more about this."
>>>
>>> You cannot force a Human readable representation in the spec. That is again a pragmatic issue. If you don't have one, then in many use cases it will be difficult for people to understand what is going on, so takeup will be slow.
>>>
>>> This will become obvious when we have more good demos.
>> okay
>>
>> 1: MUST have a representation of what exactly? just any old graph? a 
>> profile? or the public key or?
> 
> Well the next part of the protocol will make that clear. If the SPARQL query
> on that graph fails then you cannot confirm identity.
> 
>> 2: MUST have a MRD, SHOULD have a HRD || are encouraged to have an HRD - 
>> i.e. stick to only the must's or also include the very beneficial?
> 
> Well the HRD and MRD and all descriptions stuff can be moved to another section.
> If you want the howto doc, can make the point as to how it is very good practice to have an HRD (and if it does not have semantics I suggest it have at least a pointer to a machine readable doc)

agreed on both counts - good enough for me:

   "MUST have a representation in a machine readable format."
OR
   "MUST have a RDF representation."

I won't debate on which of the above two :p


More information about the foaf-protocols mailing list