Sometimes I wish for a 'debug' version of PG which adds an "order by random()" to all queries (or appends a "random()" clause to an existing "order by" clause).
Or even better: have it as a connection/session parameter.
Sufficient yes because if, for example, you’re simply sorting by a value that’s shared/duplicated like enums values, the ordering of the ordered groups is basically up to the query planner and can change (one query may use an index while another does a sequential scan, for example). The most common use case where I’ve seen this be significant is offset based pagination where the query plan actually changes depending on which page you’re on. You’ll either miss records or even get the a few rows back from the previous page because of the different ordering chosen by the planner/optimizer.
interesting! I like it.
> or appends a "random()" clause to an existing "order by" clause
That one I don't understand. If the Order By is there, why randomise it? I guess
then it sort of allows you to find out if it's been ordered sufficiently for your needs - is that it?