[postgis-users] Query with LIMIT but as random result set?

[Stefan Keller]

Hi

I have a query like this

  SELECT ST_AsText(way) geom, name AS label
  FROM osm_point
  LIMIT 10;

When I repeatedly do this, the result set will be always the same.
I have observed this only empirically and I know that the ordering of
the result set is undefined without ORDER BY.
There are two indexes involved, one geospatial for way and one for name.

My question is: Does someone have an idea on how to randomize the
result set on every consecutive query?
And as an option the (limited) resultset should be spatially
distributed (not clustered).

Yours, Stefan