Do you plan to support subqueries at some point? I’ve been trying to recreate the below.
select * from accounts where accountid in (select distinct(accountid) from accountholdings where balance > 1000)
I realize this gets very close to being a join, so whatever you plan to support around joining data would be interesting to hear. While 20megs is a lot to play with in terms of our document sizes, we may need to seperate out some of our archived data to avoid hitting this limit and so doing a join would make our lives a lot easier.
Also, will you support updates within JSON documents? For example:
Update account.holderinfo h set h.firstname=‘baz’ where h.lastname = ‘foo’