Re: "Basic profile" terminology ?

On 9 Jul 2012, at 08:46, <Erik.Wilde@emc.com> wrote:

> hello all.
> 
> On 2012-07-07 18:34 , "Henry Story" <henry.story@bblfish.net> wrote:
>> In short it is best to define the Protocol at a semantic level: you have
>> collections that can be posted into, that can be linked, that take
>> requests of documents about certain types of things... You can then
>> reason about those logically. Then one has a few well known syntaxes (
>> Turtle perhaps ) and allow content negotiation to take care of the rest.
> 
> not quite sure i am completely following what you mean here. every
> protocol is defined on some semantic level, whatever the designers of that
> protocol deem relevant and useful for the purpose of that protocol. and
> then you map the semantics to a syntax, and usually having one syntax is
> much better than having more than one syntax.

yes, but we already have a lot of different syntaxes, json, xml ( e.g., atom-xml )  
in rdf RDF/XML, Turtle, etc. There is hardly a way to impose any one standard 
format. (XML seemed to be the thing, then JSON took over. Certainly something else
will come soon ) So it is much better to make a requirement for an automatic follow your nose
method to move from syntax to semantics. ( as proposed by GRDDL for example )
Then as the syntax fashion changes general frameworks such as linked data servers
can just add a new transformation. That is easy to do with the web. As I listed
in my previous e-mail we have all the tools at our disposal: Content-Negotation,
mime types, GRDDLs, ...

> deciding on a syntax usually
> means to look at the widest possible set of potential consumers and the
> support they have for dealing with particular syntax choices, and then you
> decide based on that.

When you work in a world market every base is important and every one of these
subgroups can be millions large, and small ones can sometimes have a lot more
impact than larger ones. 


> 
> cheers,
> 
> dret.
> 

Social Web Architect
http://bblfish.net/

Received on Monday, 9 July 2012 15:37:00 UTC