RE: Issue rdfms-abouteach

>
> Issue http://www.w3.org/2000/03/rdf-tracking/#rdfms-abouteach
>
> Summary: I propose to remove rdf:aboutEach from the RDF/XML language.
>



Earlier Brian said, of a different issue:
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Oct/0362.html
>
> The barrier for making changes is set very much higher than for
> 'clarifications'.  A strong case must be made that the change is
necessary.
> There must be a strong call from the community to make the change.  And we
must
> be sure we have evidence of how existing code and data and will be
affected.
> And there must be good reason why it cannot wait till we have a more
relaxed
> charter.
>
> We should resist the temptation to fix things that we don't like but are
not
> really broken.  Lets save those for rdf 2.
>

I think we should try and apply consistent standards for changes.

I think we should explicit delete "strong call from the community" from the
necessary hurdle for changes.

However, I personally am unconvinced by Dave's case that the change is
*necessary*. aboutEach is unfortunate and shouldn't be in the language but I
feel it is in the category of "don't like but are not really broken."  (In
contrast with the issue that brought the chair into action in October where
the spec is self-contradictory and hence is broken!).

Looking at Dave's implementation overview stuff it seemed to say, aboutEach
is implementable but quite difficult and most people haven't bothered. That
is it is an implementation defect not a spec defect. The main reason for
dropping aboutEach, which isn't such a bad one, is that the corner cases to
do with aboutEach are unclear (in a range of issues raised in part by me).
The easiest way to clarify the corner cases is excision of aboutEach.
Instead of trimming five toenails, it might be easier to cut off the foot.

I suspect that laziness is the reason for not postponing this to rdf 2.

Jeremy

Received on Friday, 16 November 2001 06:26:34 UTC