Re: Bermudan flowers: Query, I18N, and syntax vs semantics

On 2002-02-04 17:17, "ext Jeremy Carroll" <jjc@hplb.hpl.hp.com> wrote:


> We could allow arbitrary typed values as node labels. e.g. in the B10
> example the nodes are labelled with the number 10.5 and not with the string
> "10.5" nor the string "10,5".

Unfortunately, Jeremy, such an approach cannot work in practice.

Otherwise all RDF parsers must support all arbitrary datatypes
and the RDF graph must provide a *lexical* representation
(albeit canonical) for all  values of all arbitrary datatypes?...


Honestly Jeremy, the situation is not as bad as you think
(though I'm sure we're not the only ones who are getting a
little glazed from the debate over these issues ;-)

Please see my comments on B10 and see if you feel afterwards
the same as you do now.


> None of the global idioms work in this view; because the global idioms are
> global, i.e. worldwide. If the global idiom interprets "10,500" as ten and a
> half thousand then it does not interpret it as ten and a half.

This is only a problem if you deviate away from TDL pairings and
try to accomodate the S idioms and semantics at the same time.

It's the fact that the S-A approach is trying to have datatypes
with mutltiple disjunct lexical spaces and mappings rather than
the single lexical space and mapping that is the root of the
problem.

If we stick to the original definition of a datatype, adopted
from XML Schema, then there is no ambiguity -- no problem.


> How about replacing the global idiom with a document scope type declaration
> idiom, possibly with an include mechanism.

Please, please, please folks. We need none of this extra
cumbersome machinery, either in the syntax or the graph.

It is only from the effort to fuse TDL and S that the
apparent (and not actual) need for such mechanisms
arises.

I think we're making this all much more difficult than it
has to be.

It has already been expressed that the MT for TDL is or
can be made sufficient -- and it does not suffer from any
of the above mentioned problems, nor from many other
problems that have come up in discussions.

I think if we would just focus on whatever tweaks are
needed for the TDL MT then we could have a good solution
and move on to other issues.

> =====
> 
> Summary of idea:
>  Use true typed values as node labels in abstract syntax.

Unfortunately, no feasible.

RDF has no native types and thus cannot have "true" values
in the graph.

We will never escape lexical forms in the graph if we are to
allow/support arbitrary datatypes.

>  Use purely syntactic means to generate these labels (built on top of
> XSD).

A bit too cumbersome, I think, especially since it can be avoided.


--

All of these supposed problems are simply arising from the
conflicts inherent in the different S idioms or between the
S idioms and the TDL idioms (or their interpretations).

I just don't see it as fruitful to keep trying to mate
animals from two different species. We're bound to keep
getting bizarre and frightening results.


Patrick


--
               
Patrick Stickler              Phone: +358 50 483 9453
Senior Research Scientist     Fax:   +358 7180 35409
Nokia Research Center         Email: patrick.stickler@nokia.com

Received on Monday, 4 February 2002 10:50:55 UTC