What are the FEATURE differences, SQL Changes, SP, Trigger, UDF changes
between DB2 Z/OS V8 and UDB 8.2?

What are the considerations for migrating from UDB 8.2 to DB2 V8 Z/OS

<db2sysc@yahoo.com> wrote in message
news:1110222601.975775.305130@g14g2000cwa.googlegr oups.com...
> What are the FEATURE differences, SQL Changes, SP, Trigger, UDF changes
> between DB2 Z/OS V8 and UDB 8.2?
>
> What are the considerations for migrating from UDB 8.2 to DB2 V8 Z/OS
>
Go to this site and download the "SQL Reference for Cross-Platform
Development":

http://www-306.ibm.com/software/data...s/v8books.html

db2sysc@yahoo.com wrote:
> What are the FEATURE differences, SQL Changes, SP, Trigger, UDF changes
> between DB2 Z/OS V8 and UDB 8.2?
>
> What are the considerations for migrating from UDB 8.2 to DB2 V8 Z/OS
>
Adhere to this:
http://www-128.ibm.com/developerwork...206sqlref.html

There si no list of differences. IBM used to publish a doc called "FRED"
which contained differences. The competition was an eager distributer....

Cheers
Serge
--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

If the competition's distribution of "the differences documentation" was so
worrying, would it not have been better to have eliminated the unnecessary
differences, rather than simply eliminating the documentation? Rather than
replacing it with a list of what's the same
(http://www.ibm.com/servers/enable/si...b2common.html), that carefully
omits things that are "wildly different"?

Mark Yudkin,
A sufferer from these unnecessary differences.

"Serge Rielau" <srielau@ca.ibm.com> wrote in message
news:393q6gF5rtolsU1@individual.net...
> db2sysc@yahoo.com wrote:
>> What are the FEATURE differences, SQL Changes, SP, Trigger, UDF changes
>> between DB2 Z/OS V8 and UDB 8.2?
>>
>> What are the considerations for migrating from UDB 8.2 to DB2 V8 Z/OS
>>
> Adhere to this:
> http://www-128.ibm.com/developerwork...206sqlref.html
>
> There si no list of differences. IBM used to publish a doc called "FRED"
> which contained differences. The competition was an eager distributer....
>
> Cheers
> Serge
> --
> Serge Rielau
> DB2 SQL Compiler Development
> IBM Toronto Lab

Mark Yudkin wrote:
> If the competition's distribution of "the differences documentation" was so
> worrying, would it not have been better to have eliminated the unnecessary
> differences, rather than simply eliminating the documentation? Rather than
> replacing it with a list of what's the same
> (http://www.ibm.com/servers/enable/si...b2common.html), that carefully
> omits things that are "wildly different"?
In the times of FRED, DB2 for LUW was in its second version, barely a
few years old. DB2 for zOS was a lot more mature.
Then again we all want our children to be born fully matured... no
hassle with pimples. ;-)

Seriously though. If you conmpare DB2 V8.2 for zOS with DB2 V8 for zOS
you will find that there has been significant convergance.
The clients is now shared as well which also helps.
Lastly the DB2 zOS folks are trapped/blessed with a lot more legacy. The
word incompatible change is a no-no in these halls, in Toronto we can at
least wisper it in private.
With DB2 for LUW there is temptation to learn from past mistakes, yet
satisfy folks like you. Darned if you don't, darned if you do.

> Mark Yudkin,
> A sufferer from these unnecessary differences.
Send me a list of these unnecessary differences and I'll see what I can
do. I thought the time of the easy pickings is long gone.

Cheers
Serge

--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

"Serge Rielau" <srielau@ca.ibm.com> wrote in message
news:395jckF5u1i2dU1@individual.net...
> Lastly the DB2 zOS folks are trapped/blessed with a lot more legacy. The
> word incompatible change is a no-no in these halls, in Toronto we can at
> least wisper it in private.
> With DB2 for LUW there is temptation to learn from past mistakes, yet
> satisfy folks like you. Darned if you don't, darned if you do.
>
> Cheers
> Serge

DB2 for LUW roots began in the OS/2 development lab in Boca Raton. They
could have made it compatible with DB2 for MVS, but they choose not to. The
folks at the OS/2 lab made a lot of "not so smart" decisions. At that time,
there was no one in charge of database architecture across IBM divisions.

If we can get the LIST OF SQL Statements which will NOT WORK in DB2 V8
Z/OS, but will be working fine in DB2 8.2 LUW, that would be great.

db2sysc@yahoo.com wrote:
> If we can get the LIST OF SQL Statements which will NOT WORK in DB2 V8
> Z/OS, but will be working fine in DB2 8.2 LUW, that would be great.
>
Anything around OO:
CREATE TYPE AS
CREATE TABLE OF
CREATE VIEW OF
METHODs

BI specific stuff:
DB2 for zOS is in the beginning w.r.t. MQTs.
OLAP functions are not supported.
MERGE statement not supported
I'm not sure how far DB2 for zOS is with CUBE and ROLLUP

DB2 for zOS does support the WITH clause. I'm not sure about recursion.

SQL PL is a bit of a leap frog game..
Not sure about nested savepoints, nested compound statement, nested
exception handlers.
DB2 for zOS is ahead in passing tables to procs
(i.e. transition tables from triggers) and in general logic in triggers
is less restricted.
SQL Functions are on par with DB2 V7.1 for LUW (RETURN statement only)
if I'm not mistaken.

I'm not sure about DGTT. DB2 for zOS supports CREATEd GTT. (i.e. cataloged)
DB2 for zOS does SELECT FROM FINAL TABLE(INSERT...), but not NEW TABLE,
OLD TABLE with UPDATE and DELETE (introduced in V8.1.4 for LUW).
But DB2 for zOS does "table-bind-in" (aka array-insert) on INSERT
statements.

That's all I can come up with ad-hoc. DB2 V8 for zOS has caught up
significantly w.r.t. generic SQL such as scalar subqueries, nesting,
etc... I think most gaps are closed. Compared to V7 there are big gaps
w.r.t. orthogonality

Cheers
Serge

PS: Note to Mark A.: I have never come across a difference that was due
to DB2 V1 and earlier. DB2 for LUW was perry much born into SQL89/SQL92,
so there is the advantage of late birth.
--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

Serge.

Thanks.

Can you explain bit more on
"DB2 for zOS is ahead in passing tables to procs
(i.e. transition tables from triggers) and in general logic in triggers

is less restricted"

any examples in passing tables?