Re: types of "graphs"

On 10-05-2012 00:30, Richard Cyganiak wrote:
> On 9 May 2012, at 22:29, Guus Schreiber wrote:
>>> Why do *we* need to define the term for a frozen g-box? Why not leave it to the community?
>>
>> Sloppy phrasing on my part. I meant to say: this line of reasoning implies there is a need for at most one type.
>
> Ok.
>
>> With my chair hat on I prefer not to take a position whether we should predefine it or not. I only note that such the "frozen-box" type  arose from our use cases.
>
> Right. If third parties can define the term without violating something in our spec text (i.e., a “frozen” g-box doesn't contradict something in our semantics), then I'd argue we should leave it to them.

[probably superfluous]

Every WG struggles with this trade-off between keeping it simple and 
providing the needed support. If we would agree that there is sound 
technical or pragmatic reason to be able to specify a graph-container 
type (but we are not there yet)  we have basically three options:

   1.  standardize it in a normative spec
   2.  suggest an (informative) best practice
   3.  do nothing

If the need for such a typing mechanism arises from many use cases there 
is certainly a case to be made for the first or second option. The 
decision might also depend on the overall landscape of proposed changes. 
But again, we're not at that stage yet.

Best,
Guus

>
> Best,
> Richard

Received on Wednesday, 9 May 2012 23:25:07 UTC