Re: G(RDDL)

Misha,

I think the problem with the idea of using RDDL to point to the
GRDDL transform is as follows.

The GRDDL architecture is that you have a finite and very small
number of things which a GRDDL client has to know and try, so that
they become definitive: data is on the SW if it is GRDDLable (or in
a sem web langauge).    Because GRDDL givges you very general hooks,
other formats may come and go, and each one can be linked to
a suitable GRDDL transform.  GRDDDL is a bootstrap pattern, in which
the minimum boot code is necessary to invoke all new systems.
It is a question of the minimum amount of knowledge which will open
the door to everything on the web. It includes a couple of places to  
look for pointers and XSLT.

You seem to be arguing that RDDL should have GRDDL roles.
Now if RDDL were to be part of the GRDDL in that way, it would
have to be part of the GRDDL boot sequence.  A GRDDL engine would
have to understand those roles to use them. The whole point about
GRDDL is that it is used by agents with no prior knowledge of the  
format.
So you would need all GRDDL systems to understand the RDDL spec.
You add the RDDL spec to the GRDDDL boot core knowledge.

But what does a GRDDL system gain from that?  Only a knowledge of
how to extract data from RDDL, as the namespace document is RDDL.
Which it knew anyway, because it understood the RDDL spec. So the  
gain is zero.

What does work is to put in the RDDL namespace document a GRDDL pointer,
so that GRDDL-aware systems can slurp up RDDL -- without any prior  
knowledge of
RDDL.

Tim



On Mar 9, 2006, at 10:22, Misha Wolf wrote:

>
> In seeking to formulate a proposal to the NewsML 2 Architecture WP
> for how we should make use of RDDL and GRDDL, I am struck by the
> different approaches taken by these two specs to linking to
> resources.  Presumably, there is no reason why not to use the RDDL
> approach to link to GRDDL transforms (instead of the approach
> proposed in the GRDDL spec).  If we were to do this, we would wish
> to use some agreed natures/purposes.  Is anyone interested in
> defining these?
>
> Thanks,
> Misha
>
> NewsML 2 resources:
> http://www.iptc.org/        | http://www.iptc.org/NAR/
> http://www.iptc.org/NAR/1.0 | http://groups.yahoo.com/group/newsml-2/
>
>
> To find out more about Reuters visit www.about.reuters.com
>
> Any views expressed in this message are those of the individual  
> sender, except where the sender specifically states them to be the  
> views of Reuters Ltd.
>

Received on Saturday, 11 March 2006 01:11:26 UTC