Re: Restpark - Minimal RESTful API for querying RDF triples

On 4/16/13 3:28 PM, Aidan Hogan wrote:
>
> In practice, it is not at all difficult to find (interesting) SPARQL 
> queries that existing engines cannot run completely and correctly. I 
> (unfortunately) find them all the time without even trying.
>
> SPARQL evaluation does not scale, cannot scale, will never scale.

That should depend on the combination of query, dataset size, computing 
power, and SPARQL engine quality. In the worst case, the query should 
simple conclude: I cannot provide a solution within the time allotted, 
so give me more time....

I would be very happy to look at some of those queries thrown at the 
instance we maintain at: http://lod.openlinksw.com . It has a massive 
data collection (50+ billion triples) and a publicly accessible SPARQL 
endpoint. As I said, your worst case should be: give me more time to 
compute a solution or the query optimizer has determined that your 
solution simply requires too much computing resources etc..

Back to you, send me SPARQL queries or SPARQL protocol URLs.

>
> But maybe I misunderstand. What do you mean when you say "SPARQL 
> scales"? :)

There's one SPARQL query language and wire protocol combination. Not so 
re. any RESTful interface since there will never be consensus which 
means its a steady march to a zillion interaction patterns and media 
types etc..
>
>
> With respect to Restpark, binary search is logarithmic and answering a 
> Restpark query would require one or two lookups. Some potential 
> problems for scale arise from low-selectivity lookups, but that's a 
> drop in the ocean compared to what's faced by SPARQL engines. 


-- 

Regards,

Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen