Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>> [implement "clustered index" as a covering index with all columns
>> which are present in the heap]
> I guess we could do that, but I'm not convinced there would be
> much benefit.
The "traditional" way to implement a clustered index is to have the
leaf level of the index contain the tuples rather than pointers to
the tuples. If we're going to do clustered tables, we might want to
jump all the way to that, rather than a half-way solution which
stores everything twice.
-Kevin