Re: allow implicitly unbound variables in SPARQL results? (formatting fixed)

Kendall Clark <kendall@monkeyfist.com> wrote on 12/14/2005 12:02:51 PM:

> 
> On Dec 14, 2005, at 11:27 AM, Lee Feigenbaum wrote:
> 
> >
> > 'c' eliminates the indexed-based approach to parsing the results. At
> > the very least, I feel that if the Working Group decides on option 'c'
> > than the requirement that variable name order be preserved within
> > the resultset should be removed.
> 
> I don't think UMD would object to that. I've never thought that was 
> worth the pain anyway.
> 

To reply to two points at once; you asked for clarification of in what way 
the collapsed option is not fully-functional. It is not fully functional 
in that it does not support position-based indexing as the LC design does. 
If that is not a desirable property, then I believe that the best approach 
is to eliminate this requirement from the results format (and then going 
with option 'c' is less objectionable).

Lee