Adrian Klaver schrieb am 15.01.2019 um 17:44:
So we end up with something like this:
select distinct on (C.contact_id) C.contact_id, C.lname, C.fname,
C.direct_phone, O.org_name, A.next_contact
from Contacts as C
join Organizations as O on C.org_id = O.org_id
join Activities as A on C.contact_id = A.contact_id
where A.next_contact <= 'today'
and A.next_contact > '2018-12-31'
order by c.contact_id, a.next_contact DESC;
And I've never liked this method (though I'm old and crotchety)....
What is the specific objection?
To me:
1) Plus side. It saves an explicit sub-query
2) Minus side. It is not portable. But then complete portability is a unicorn
in my opinion.
I think Ron was referring to the JOIN operator, rather than the DISTINCT ON
