Re: not sure if we're done with: Example Errors (sections 8, 9, 10)

On Tue, Nov 22, 2005 at 07:36:28AM -0500, Eric Prud'hommeaux wrote:
> On Tue, Nov 22, 2005 at 11:45:19AM +0000, Seaborne, Andy wrote:
> > 
> > 
> > 
> > Dan Connolly wrote:
> > >Andy, you wrote:
> > >
> > >"... There wil be a formal response to your email 
> > >when all the comments have been acted on."
> > >
> > >and I'm not sure if anything relevant has happened since then.
> > >Any clues?
> > >
> > 
> > All the edit changes (sec 8/9/10) are done from:
> > 
> > http://lists.w3.org/Archives/Public/public-rdf-dawg-comments/2005Sep/0011.html
> > 
> > Eric - what's the status of the comments on sec 11?
> 
> 
> The first example in 11.2.3.1 for sop:RDFterm-equal should have two results
> in the query results section, with the inverse solution as well as the
> listed, since name1=>Ms A. and name2=>Alice is just as valid a binding.
> 
> fixed
>   http://www.w3.org/2001/sw/DataAccess/rq23/#conformanceB
> 
> The second example in 11.2.3.1 has a typo, the time in the data should be
> 19:00, not 19:01 like listed, to get the results expected.
> 
> fixed
>   http://unagi/2001/sw/DataAccess/rq23/#func-RDFterm-equal
> 
> The first example in 11.2.3.2 has several problems, so it's best to
> re-examine it.  Namely, the FILTER construct is missing a parenthesis (or
> should have the outside one removed) and the example appears to be
> projecting an undefined variable for no reason.
> 
> fixed
>   http://unagi/2001/sw/DataAccess/rq23/#func-isBound
> 
> The second example in 11.2.3.2 should have foaf:givenName as the predicate,
> not foaf:name.
> 
> aha, tx
> fixed
>   http://unagi/2001/sw/DataAccess/rq23/#func-isBound
> 
> The query under section 11 Testing Values should probably read "?date > ..."
> and not "?date < ...".  Otherwise, the same result will be returned in both
> examples, since neither of the dates in the data will test true, no matter
> whether the casting works or not.
> 
> fixed
>   http://www.w3.org/2001/sw/DataAccess/rq23/#tests
> 
> I note that when I ran the LC version, I got no results.
> 
> The example in 11.2.3.6 should read "FILTER regex(?name...", not "FILTER
> regex(name..."
> 
> fixed
>   http://www.w3.org/2001/sw/DataAccess/rq23/#funcex-regex
> 
> This may actually present some issues in general, but example 11.2.3.8 has a
> statement ""FILTER (lang(?name) = "ES" ) )"".  Technically, "ES" will parse
> to a untyped literal and the LANG function returns a xsd:string, which will
> by the constraints of the language result in a literal compare, which will
> fail in all cases.  To be technically accurate, it should read ""FILTER
> (lang(?name) = str("ES") ) )"".

Yes, that was the product of excessively sprinkling xsd types where
they weren't needed. I have updated this text to reflect my 
implementation and several others (note the announcement and responses
[ST]). The editor's draft now has a new type called "simple literal",
defined as an RDF plain literal with no language tag. The operators
STR and LANG return simple literals; the operators <, >, <=, >=,
langMATCHES and REGEX have forms that take simple literals as
arguments.

The examples in the spec reflected the actul implementations. They
have not changed. The operator definitions should now be consistent
with the examples (well, they *should* have been consistent before,
but now I think they are).



Andy, does this complete the response to Example Errors? Let me know
if you need anything else. 

[ST] http://lists.w3.org/Archives/Public/public-rdf-dawg/2005OctDec/thread#msg304
-- 
-eric

office: +81.466.49.1170 W3C, Keio Research Institute at SFC,
                        Shonan Fujisawa Campus, Keio University,
                        5322 Endo, Fujisawa, Kanagawa 252-8520
                        JAPAN
        +1.617.258.5741 NE43-344, MIT, Cambridge, MA 02144 USA
cell:   +81.90.6533.3882

(eric@w3.org)
Feel free to forward this message to any list for any purpose other than
email address distribution.