I'm implementing paging through search results using cursors. Is there a
better way to know total number of rows under a cursor than running a
separate COUNT(*) query? I think PostgreSQL is bound to know this number
after the first FETCH, isn't it?
On a side note, why queries using LIMIT are SO terribly slow, compared to
cursors and sometimes even ones without LIMIT? Shouldn't LIMIT be internally
implemented using cursor mechanism then?
Best regards,
Egor Shipovalov.