I really would like to see native result set caching sooner than
later. Any plans for such opulence? :-)
Plans? yes ... of a sort!
I have been toying with ideas of front-ending a SPARQL endpoint with
query cache for repeats. Web caching only helps at long range;
applications do seem to issue the same Q
Is that what you meant?
I also thought it could optimizer "paging" requests i.e. look for
queries with ORDER BY-OFFSET-LIMIT in the expectation a query with
OFFSET+N is about to arrive. Might as well execute without OFFSET-LIMIT
once, then slice the cached results.
Couple with version numbering in TxTDB as transactions pass through and
cache maintenance can be done with ETags and conditional GETs.
Andy
