Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Weak/strong typing in DB2 V9.7

26 views
Skip to first unread message

Damir

unread,
Dec 1, 2009, 6:50:18 AM12/1/09
to
Hi,
is there a registry variable (or DB parameter, or something else) that
enables/disables weak typing in DB2 V9.7?
Cannot find it in the docs...
Or is it (weak typing) turned on by default (yes it is, AFAIK) and cannot be
disabled?
Thanks,
Damir

Serge Rielau

unread,
Dec 1, 2009, 8:11:35 AM12/1/09
to
It's a brave new world....no switch.
It's too deeply integrated to have a switch for it.
Do you want to switch it of generally or are there specific scenario's
you'd like to control?

Cheers
Serge

--
Serge Rielau
SQL Architect DB2 for LUW
IBM Toronto Lab

darko

unread,
Dec 1, 2009, 1:02:11 PM12/1/09
to

How do you feel about that, Serge? :-)

Darko

Serge Rielau

unread,
Dec 1, 2009, 2:05:53 PM12/1/09
to
> How do you feel about that, Serge? :-)
I hate the feature. How's that for an answer?
Unfortunately DB2 is used in the real world and has to respond to real
industry pressures.
My wishes are secondary.

darko

unread,
Dec 1, 2009, 5:06:14 PM12/1/09
to

Points for your honesty :-)

I am ambivalent personally, but our developers would perhaps force me
to move to 9.7 immediately just for the sake of weak typing, if they
knew it is available. I had hard moments very often defending DB2
against Informix or MySQL, being asked: "Why can't DB2 understand what
I need, like Informix does? I have to type much more!" and such
questions. My unfair, but most efficient quick counter-argument was
often: "Do you want to have a RDBMS that lets you define column as
number, then insert a character/string in it without error, like some
of them does?" (not Informix, AFAIK). More serious argument I have
used was that the trouble with computers generally is that they
usually do what you order them to do, not what you want them to
do :-)

When mentioning 9.7, is it safe to use it for production OLTP and DWE?
Or to settle with 9.5 for some time? We are about to migrate from 8.2.

Darko Krstic

Serge Rielau

unread,
Dec 1, 2009, 6:05:02 PM12/1/09
to
darko wrote:
> I am ambivalent personally, but our developers would perhaps force me
> to move to 9.7 immediately just for the sake of weak typing, if they
> knew it is available. I had hard moments very often defending DB2
> against Informix or MySQL, being asked: "Why can't DB2 understand what
> I need, like Informix does? I have to type much more!" and such
> questions.
There, you make my point. If your developers perceived DB2 as "inferior"
because it insisted on correct typing. That is not a winning position -
only a righteous one.

> My unfair, but most efficient quick counter-argument was
> often: "Do you want to have a RDBMS that lets you define column as
> number, then insert a character/string in it without error, like some
> of them does?" (not Informix, AFAIK). More serious argument I have
> used was that the trouble with computers generally is that they
> usually do what you order them to do, not what you want them to
> do :-)

I spent two hours today fighting a bad performing query until finally
thinking of comparing join member types.
Also today there was a post on BOOLEAN in Oracle on the same topic.

> When mentioning 9.7, is it safe to use it for production OLTP and DWE?
> Or to settle with 9.5 for some time? We are about to migrate from 8.2.

We have shipped fixpack 1. It's safe and very valuable.
* Biggest delivery of SQL and SQL PL since DB2 V2
* Triple the coverage for compression (temp + index)
* Much improved monitoring infrastructure
* ability for existing databases to move into automatic storage over time
* Kissing good bye the the high watermark!
* Greatly simplified schema evolution
* "Readers don't block writers and writers don't block readers"

Well those are just my favorites

Damir

unread,
Dec 2, 2009, 5:26:15 AM12/2/09
to
Couldn't agree more with both of you!
We would like to have this switch because we feel strong typing is safer (as
in "less error prone"), but then the "ease-of-use" and "less-tedious"
arguments kick in from the other side (lazy programming, or lazy
programmers?!), so we will live happily without it :-)
We are now in test with 9.7.1 and since all is (looking) well (stable) we
have plans to go to the production as well, right after the new year.
My personal favourite: partitioned indexes (compression comes in as extra
bonus).

Regards,
Damir

0 new messages