Column aliases for GROUP BY and HAVING

(Postgres version 8.2.4)

Trying to understand GROUP BY, I'm reading on
<http://www.postgresql.org/docs/8.4/interactive/queries-table-expressions.html>.
Consider this query:

=> select x, sum(y) from test1 group by x order by x;
 x | sum
---+-----
 a |   4
 b |   5
 c |   2
(3 rows)

I understand what's happening here, all is fine. So I play with HAVING:

=> select x, sum(y) from test1 group by x having sum(y)>3 order by x;
 x | sum
---+-----
 a |   4
 b |   5
(2 rows)


But this confuses me:

=> select x, sum(y) as foo from test1 group by x having foo>3 order by x;
ERROR:  column "foo" does not exist
LINE 1: ...ect x, sum(y) as foo from test1 group by x having foo>3 orde...

Why isn't it possible to refer to a column alias in HAVING?


--
- Rikard