proposed rescindment of errata-query-10

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

proposed rescindment of errata-query-10

Peter F. Patel-Schneider
errata-query-10 does not suggest how to change the SPARQL 1.1 Query
Language specification.

The referenced discussion is no help, as it suggests only changing
occurences in basic graph patterns, which would actually make EXISTS work
worse than it currently does, as substitution for ?n would not happen in the
inner FILTER in the example in Section 8.3.3

PREFIX : <http://example.com/>
SELECT * WHERE {
        ?x :p ?n
        FILTER NOT EXISTS {
                ?x :q ?m .
                FILTER(?n = ?m)
        }
}

This leads to results counter to the discussion around this example.

Both suggestions in erratum are unspecific.  The suggestion in the
referenced discussion does not propose a workable solution.

The erratum should be rescinded until a suitable replacement can be crafted.


Peter F. Patel-Schneider
Nuance Communications



Reply | Threaded
Open this post in threaded view
|

Re: proposed rescindment of errata-query-10

Andy Seaborne-4
On 23/06/16 23:20, Peter F. Patel-Schneider wrote:
> errata-query-10 does not suggest how to change the SPARQL 1.1 Query
> Language specification.

One purpose of the errata document is to capture issues for a future
working group. The criteria for inclusion is as a suggestion something
needs looking at in the next WG, not that that a detailed fix is available.

    Andy


>
> The referenced discussion is no help, as it suggests only changing
> occurences in basic graph patterns, which would actually make EXISTS work
> worse than it currently does, as substitution for ?n would not happen in the
> inner FILTER in the example in Section 8.3.3
>
> PREFIX : <http://example.com/>
> SELECT * WHERE {
>          ?x :p ?n
>          FILTER NOT EXISTS {
>                  ?x :q ?m .
>                  FILTER(?n = ?m)
>          }
> }
>
> This leads to results counter to the discussion around this example.
>
> Both suggestions in erratum are unspecific.  The suggestion in the
> referenced discussion does not propose a workable solution.
>
> The erratum should be rescinded until a suitable replacement can be crafted.
>
>
> Peter F. Patel-Schneider
> Nuance Communications
>
>
>