Florian G. Pflug wrote:
> knowing *why* the standard committee disallows that syntax -
> and why everybody except oracle chose to agree with it would be quite
> interesting.
Consider that we rightfully disallow
SELECT * FROM tab1, tab1;
because there are two tables with the same name involved. This hints to
the fact that the internal query processing needs to have a unique name
for each table source. You could argue why that would have to be, but
it seems quite useful. Otherwise you would also have to allow
SELECT * FROM tab1 AS a, tab2 AS a;
and that cannot possibly be good.
If you believe that, then a table source without any name at all, such
as one produced by subqueries, should also be disallowed. Requiring a
unique name appears to be a stronger condition than requiring a name at
all. This is actually false if you consider that SQL itself does
distinguish between unique constraints and not-null constraints. But
you don't really want to invent a logic by which nameless table sources
are managed. Strict logic would probably require the results of all
such queries to be unknown.
Note that Oracle is notorious for mishandling null values in various
ways. So it's not surprising that they mishandle "null" table names in
what appears to be pretty much the same way.
--
Peter Eisentraut
http://developer.postgresql.org/~petere/