* Ben (bench(at)silentmedia(dot)com) wrote:
> Oracle has a very handy function called first_value, which can be used
> to turn a set like this:
>
> a 10
> a 3
> b 1
> c 30
> c 10
> d 1
> d 20
>
> ...into this:
>
> a 10
> b 1
> c 30
> d 1
>
> Does postgres have something equivalent, or, even better, is there a
> reasonable way to express this in standard SQL?
Isn't this just: distinct on (col1)?
Stephen