ext Philip Van Hoof <[email protected]> writes: >>> Nothing prevents from having 2 DB, [...] >> >> Also, we've had > 1 DB before in Tracker [...] > > Some technical chitchat on this:
I think Igor is just proposing that we do the indexing lazily (not counting it towards the import operation itself), and somehow notify the user that indexing is still ongoing or give him/her the option to do a simple grep on the non-indexed part. Anyone up for writing a SPARQL query engine that swallows TTL files? We could then do batch imports by giving TTL files to Tracker, index them lazily, and while there still is a non-indexed TTL file around, we run the slow SPARQL TTL query engine on them. Does that make any sense? _______________________________________________ MeeGo-dev mailing list [email protected] http://lists.meego.com/listinfo/meego-dev http://wiki.meego.com/Mailing_list_guidelines
