
10 Oct
2012
10 Oct
'12
midnight
Toby Corkindale <toby.corkindale@strategicdata.com.au> wrote:
However, I think you're missing a point. Sometimes you are really storing a whole document, or at least a large amount of very structured, self-contained data. For these cases, JSON storage is quite useful and document storage is a lot neater (and performant) than having to write a whole heap of code to deparse and reparse it into all the constituent components.
As I recall, this is where database systems differ from information retrieval systems. The latter are designed to store documents (or at least document fragments). Their search strategies and result ranking algorithms are designed accordingly.