On Jun 15, 7:36 pm, Serge Rielau <srie...@ca.ibm.com> wrote:
> Hi folks,
>
> My fellow team mates had some extra time on their hands so we decided to
> spice up DB2 with a grab-bag of compatibility features.
> We wouldn't mind help validating the semantics match....
> So for those of you blessed or cursed with a competitive DBMS here is a
> list of what we've added:
>
> 1. ** DUAL
> Always works without prefixing a schema
> 2. ** ROWNUM
> 3. ** (+) outer join syntax
> 4. LEAST/GREATEST/NVL/DECODE
> 5. TO_DATE/TO_CHAR improvement
> DB2 supports most common patterns except those requiring language
> awareness
> 6. ** CONNECT BY
> This is a function drop, performance drop will follow
> 7. A slew of syntactic sugar like:
> Seq.NEXTVAL and seq.CURRVAL notation
> UNIQUE instead of DISTINCT
> MINUS instead of EXCEPT
> Unnamed nested subqueries (aka inline views)
> "SELECT * FROM (SELECT * FROM T)"
> CROSSJOIN
> 8. BITAND/BITOR/.....
>
> The features marked with ** require a registry setting:
> db2set DB2_COMPATIBILITY_VECTOR=3F
> should switch everything on.
>
> There are other features those porting apps will value:
> GLOBAL VARIABLES
> A new ARRAY data type
> A new RID() function can be used to map ROWID
>
> Docs:https://publib.boulder.ibm.com/infoc...v9r5/index.jsp
>
> Enjoy
> Serge
>
> --
> Serge Rielau
> DB2 Solutions Development
> IBM Toronto Lab
>
> --
> Serge Rielau
> DB2 Solutions Development
> IBM Toronto Lab

What about:

"Row-comparisons (Bernard's favorite)
CREATE TABLE T1(c1 int, c2 int);
SELECT * FROM T1 WHERE (c1, c2) > (?, ?)
"

Bernard (Dhooghe)