Re: Expressions in SELECT

On 19 Oct 2006, at 00:17, Lee Feigenbaum wrote:

>
> Andy Seaborne wrote on 10/17/2006 01:45:17 PM:
>
> > Fred,
> >
> > The design message I metioned was:
> >
> > http://lists.w3.org/Archives/Public/public-rdf-dawg/2006AprJun/0145
> >
> > An explicit assigment might be the best long term approach if it  
> is also
> > available in the main part of the query, otherwise a syntax that is
> > specific to the SELECT clause seems more appropriate.  To be  
> compatible
> > with the results format as it stands, there needs to be a  
> variable name
> > - that could change to a more general label - but just positional  
> does
> > not work because the XML results format omits unbound variables, not
> > record them explicitly.
>
> To hopefully help hasten a decision on the punctuationSyntax issue,  
> I just wanted to write a note to state that we're (Elias and I --  
> IBM) in favor of keeping the SELECT clause syntax as-is for the  
> current version of SPARQL. We're sympathetic to people's desires to  
> allow expressions in the SELECT list, but are also confident (in  
> part thanks to Andy's message referenced above) that keeping the  
> current comma-less design will not prohibit expressions in the  
> SELECT list in the future.

All other things being equal I would prefer to have mandatory commas  
in all lists, but Garlik has a considerable number of queries backing  
our system that work without commas, and is in favour of a speedy  
resolution. So, we are also in favour of keeping the SELECT syntax as- 
is.

- Steve