Serge Rielau wrote:

FETCH FIRST n ROWS ONLY in query expressions (URC-032) (IBM lead)

And OFFSET, it seems.

TRUNCATE TABLE (E3F-012) (IBM lead)

Hopefully, TRUNCATE TABLE cannot be stated unless it's the only operation
in the transaction? (I find the SQL:2008 document hard to interpret
regarding this.)

"Who cares?"

[...]

I'm very happy that the standard is alive and kicking. MERGE and
recursive SQL are good examples of areas where the standard is certainly
not being ignored.

Among the Big Three I've found MSSQL to be the fastest moving product
regarding compliance improvements. Oracle seems to be the slowest, by far.

It would be great with sort of a road map regarding DB2: When will the
collation support be improved? What about datetime features?

--
Regards,
Troels Arvin <troels@arvin.dk>
http://troels.arvin.dk/

Troels Arvin wrote:

Serge Rielau wrote:

And OFFSET, it seems.

Yes. We thought it's about high time teh standard says anything on teh
matter. This has been the wild wild west for too long.

>
Hopefully, TRUNCATE TABLE cannot be stated unless it's the only operation
in the transaction? (I find the SQL:2008 document hard to interpret
regarding this.)

You mean unless it can be rolled back...

I'm very happy that the standard is alive and kicking. MERGE and
recursive SQL are good examples of areas where the standard is certainly
not being ignored.

Yes, I was pleased by what I saw in MS SQL Server 2008 on MERGE.

Among the Big Three I've found MSSQL to be the fastest moving product
regarding compliance improvements.

Uhm.. If you come from that far away it's easy to improve a lot ;-)

Oracle seems to be the slowest, by far.

Of course. If you are the biggest fish in the distributed pond customer
lock-in is a really attractive option.

It would be great with sort of a road map regarding DB2: When will the
collation support be improved? What about datetime features?

*Hmmhmmmhmmmhmmmmmbll*

Cheers
Serge
--
Serge Rielau
DB2 Solutions Development
IBM Toronto Lab