database market share 2003

Jean-David Beyer wrote:

nobody wrote:
Its pathetic that you can't write good code. ;-)

It's pathetic that you top post. ;-)

If you can grok C++, then you should be able to write thread safe C
code and then use them.

Well, this flame war was unexpected.

Do people still disagree that rkuserthingy is a dickhead for starting this
thing?

Nov 12 '05 #153

Serge Rielau wrote:

Even if extenders would need to be changed (which I don't know) when
going against DB2 AS/400 or DB2 z/Series it would still be irrelevant
since Oracle cartridges are effectively a no show on these platforms.

Cheers
Serge

Oracle 'cartridges'? How long as it been since you've looked at Oracle?

--
Daniel Morgan
http://www.outreach.washington.edu/e...ad/oad_crs.asp
http://www.outreach.washington.edu/e...oa/aoa_crs.asp
da******@x.washington.edu
(replace 'x' with a 'u' to reply)

Nov 12 '05 #154

Serge Rielau wrote:

Noons wrote:
What was stated here many times by IBM people was that the
code base was the same. Unqualified. Period. Whereas
Oracle's wasn't. About as much sense as the rest of
IBM's policies...

Noons, can you point us to a google-link where this claim was made. I'm
following these newsgroups since about 6 years and I have never seen
such a claim.

Cheers
Serge

I just went to google and put in the following as my seach criterion:

"DB2" AND "Single code base"

You might want to do the same.

The first link returned was:
www-1.ibm.com/....
Top of the next page was:
www.db2mag.com/ story/....
And Oracle ... at otn.oracle.com has White Papers
detailing where that claim has been made.

I think the claim has been made numerous times. And like so much
marketing hyperbole is one that would best be apologized for.
--
Daniel Morgan
http://www.outreach.washington.edu/e...ad/oad_crs.asp
http://www.outreach.washington.edu/e...oa/aoa_crs.asp
da******@x.washington.edu
(replace 'x' with a 'u' to reply)

Nov 12 '05 #155

Noons

Daniel Morgan allegedly said,on my timestamp of 9/06/2004 3:30 PM:

I think the claim has been made numerous times. And like so much
marketing hyperbole is one that would best be apologized for.

Yeah, but let's be fair: Serge was talking about a specific thread
in the newsgroups where I and a few IBMers got into a bunfight
for a while. Pity it is not complete in google. Blair has promised
to send me his posts so I can try to make sense of what was
said back then by so many of us.

We all know marketeers were AWOL the day God was handing out
intelligence. I think it goes for all sides equally: Oracle
has come out with some doozies over the years. Let's not mention
the other m$ob...

The whole single code base thing is quite bogus, IMHO. Just
exactly when is a product single code base? I mean, in my time
I've written a fair amount of #define and #ifdef pairs to try
and make things work properly everywhere, and that was JUST in the
UNIX environment. You should have seen some of my DOS/VSE Assembler
macros! I wouldn't recommend that type of coding to anyone, but
it had to be done to get over the problems way back then.

In a way the Java generation has it easy: they hardly ever need
to bother with this crap. On the other hand they have to contend
with J2EE and the lesser said about that, the better...

--
Cheers
Nuno Souto
wi*******@yahoo.com.au.nospam

Nov 12 '05 #156

Larry

And I suppose that Oracle has no cross to bear in this area?

Larry Edelstein

Daniel Morgan wrote:

Serge Rielau wrote:
Noons wrote:
What was stated here many times by IBM people was that the
code base was the same. Unqualified. Period. Whereas
Oracle's wasn't. About as much sense as the rest of
IBM's policies...

Noons, can you point us to a google-link where this claim was made.
I'm following these newsgroups since about 6 years and I have never
seen such a claim.

Cheers
Serge

I just went to google and put in the following as my seach criterion:

"DB2" AND "Single code base"

You might want to do the same.

The first link returned was:
www-1.ibm.com/....
Top of the next page was:
www.db2mag.com/ story/....
And Oracle ... at otn.oracle.com has White Papers
detailing where that claim has been made.

I think the claim has been made numerous times. And like so much
marketing hyperbole is one that would best be apologized for.

Nov 12 '05 #157

Jean-David Beyer

Noons wrote (in part):

The whole single code base thing is quite bogus, IMHO. Just exactly
when is a product single code base? I mean, in my time I've written a
fair amount of #define and #ifdef pairs to try and make things work
properly everywhere, and that was JUST in the UNIX environment. You
should have seen some of my DOS/VSE Assembler macros! I wouldn't
recommend that type of coding to anyone, but it had to be done to get
over the problems way back then.

I used to have to write portable (among UNIX systems) code and as the
number of systems went up, the number of #ifdef #endif pairs went up as
well, to the point that code became unreadable (even though it worked).
Since we used makefiles anyway, we changed things around so that the
differences were mainly in header files instead of the #ifdef #endif
pairs. And the makefiles ensured the correct header files were included.

The stuff I was working with was related to assembly-level optimizers, so
that was not as easily done. In that case, the makefile determined from
the target what source files to use as the code for an AT&T 32100 chipset
differed too much from an Intel 80386 from a Motorola 68030 from a SPARC
from a MIPS ...: you just could not use the same assembly level optimizer
for each. But there was a lot of common stuff even so.
--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ Registered Machine 241939.
/( )\ Shrewsbury, New Jersey http://counter.li.org
^^-^^ 09:05:00 up 1 day, 18:30, 4 users, load average: 4.20, 4.12, 4.12

Nov 12 '05 #158

You mean this:
http://www-1.ibm.com/industries/fina...270311103.html
"* Multiple operating systems, running on IBM and non-IBM hardware
* A single code base to deliver a consistent set of functionality
from small single CPU systems, to SMP, to clusters of either"

That statement is correct.

One can pick on the fact that the author didn't fully qualify "DB2 UDB"
to "DB2 UDB for Linux Unix and Windows".
We discussed this "legacy" problem of the UDB qualifier in earlier posts
in this thread. UDB is unfortunately commonly used similarly to "SQL".
"SQL" means MS SQL Server, UDB means DB2 UDB for Linux Unix and Windows.

I recall a debate with Mark on an Oracle claim alleging that "inline
procedural language" wasn't new and Oracle has it. After explaining to
Mark over a beer what that feature means we settled for "marketing
doesn't know better".
Not every fine slip is malicious.

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

Nov 12 '05 #159

Daniel Morgan wrote:

Serge Rielau wrote:
Even if extenders would need to be changed (which I don't know) when
going against DB2 AS/400 or DB2 z/Series it would still be irrelevant
since Oracle cartridges are effectively a no show on these platforms.

Cheers
Serge

Oracle 'cartridges'? How long as it been since you've looked at Oracle?

Well, everyone knows what I meant, right? What is the name du jour? Package?

--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

Nov 12 '05 #160

Noons

Serge Rielau allegedly said,on my timestamp of 9/06/2004 11:49 PM:

One can pick on the fact that the author didn't fully qualify "DB2 UDB"
to "DB2 UDB for Linux Unix and Windows".
We discussed this "legacy" problem of the UDB qualifier in earlier posts
in this thread. UDB is unfortunately commonly used similarly to "SQL".
"SQL" means MS SQL Server, UDB means DB2 UDB for Linux Unix and Windows.
The whole thing could be dropped if IBM used the terms it has always used:
AS400 SQL, mainframe DB2 and UDB for the rest.
This idea that "everything is DB2" but in fact isn't is just
confusing, open to mis-interpretation and a dirt easy target.
Whoever came up with this at IBM should be summarily shot...
Mark over a beer what that feature means we settled for "marketing
doesn't know better".

Yup, I'd go along with that.

--
Cheers
Nuno Souto
wi*******@yahoo.com.au.nospam

Nov 12 '05 #161

Blair Adamache

Noons wrote:

.... In a way the Java generation has it easy: they hardly ever need
to bother with this crap...

Except that using a common JDK level on any range of operating systems
is almost impossible - if you're 32- and 64-bit, you end up supporting
JDK 1.3.n in some places and 1.4.0 and 1.4.1+ in others.

Nov 12 '05 #162

Noons

Blair Adamache allegedly said,on my timestamp of 10/06/2004 12:29 AM:

Except that using a common JDK level on any range of operating systems
is almost impossible - if you're 32- and 64-bit, you end up supporting
JDK 1.3.n in some places and 1.4.0 and 1.4.1+ in others.

Oh no! I forgot about that one, it's gonna be fun and
games to move everything to 64: gonna take years.
--
Cheers
Nuno Souto
wi*******@yahoo.com.au.nospam

Nov 12 '05 #163

Noons wrote:

The whole thing could be dropped if IBM used the terms it has always used:
AS400 SQL, mainframe DB2 and UDB for the rest.
This idea that "everything is DB2" but in fact isn't is just
confusing, open to mis-interpretation and a dirt easy target.
Whoever came up with this at IBM should be summarily shot...

I disagree. Everyone equates Oracle (no qualifier) with the Relational
database, yet the same name is used for Apps, and of course the mobile
database offering.
IMHO the problem is that DB2 "for the rest of us" never got a clean name
that kept steady for several releases.
The lastest crime is "for Muliplatforms" which sounds like Milla
Jovovich in the "5. Element" and is largely ignored.
And of course there is UWO, LUW, "distributed platforms", "common
server", ...
We had similar problems with "SQL Procedural Langage" (SQL PL) which
isn't a language at all, really.. It's just SQL.
Things need short memorisable names, the market demands it, and if there
is none chaos prevails.
It's an IBM blindspot.

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

Nov 12 '05 #164

Blair Adamache

You can't gather facts by searching on google, looking at the first two
lines, and reaching a conclusion. It's a good way to stuff one's head
with assumptions and hyperbole, but you have to read the links you find
to start learning anything. Good God, they must teach research
methodologies at the University of Washington, right?

And please let everyone refrain from searches like "Blair
Adamache"+wanker or the same for Daniel, Serge or Nuno.

Daniel Morgan wrote:

I just went to google and put in the following as my seach criterion:

"DB2" AND "Single code base"

You might want to do the same.

The first link returned was:
www-1.ibm.com/....
Top of the next page was:
www.db2mag.com/ story/....
And Oracle ... at otn.oracle.com has White Papers
detailing where that claim has been made.

I think the claim has been made numerous times. And like so much
marketing hyperbole is one that would best be apologized for.

Nov 12 '05 #165

nobody

Actually, Oracle does a lousy job of extesibility.
Informix has it done right, however, it took them quite some time in
getting Stonebraker's stuff morphed in to IDS.

Its a pitty cause that's what helped cause Phil's downfall. He saw the
vision, yet no one understood how to sell/market it and how to integrate
it in to the existing platform.

I worked with Oracle 8i. Their "object relational" approach was garbage
then and I seriously doubt things have changed much.
Serge Rielau wrote:

Daniel Morgan wrote:
Serge Rielau wrote:
Even if extenders would need to be changed (which I don't know) when
going against DB2 AS/400 or DB2 z/Series it would still be irrelevant
since Oracle cartridges are effectively a no show on these platforms.

Cheers
Serge

Oracle 'cartridges'? How long as it been since you've looked at Oracle?

Well, everyone knows what I meant, right? What is the name du jour?
Package?

Nov 12 '05 #166

"SQL" means MS SQL Server
Something I'm sure MS are more than happy with.
UDB means DB2 UDB for Linux Unix and Windows.

This I just don't buy. You say this, but the IBM web site refers to
DB2 Universal Database for .... when referring to all three code bases.

And you have documents entitled "DB2 UDB Family On Common Ground".

And these documents say things like "In February of 1999, IBM® announced
DB2® Universal Database for iSeries (DB2 UDB for iSeries), ..."

Searching for all the following keywords on the IBM website returns vast
numbers of hits
DB2 UDB iSeries - 8,691 hits
DB2 UDB OS/390 - 8,095 hits
Note that the total nbr of hits for DB2 UDB is only 56,445
This is not accidental or even the result of a legacy. I firmly believe
that somebody somewhere at IBM is more than happy that some people read
UDB and mistakenly believe that it's all the same product.

Otherwise all this stuff would have been corrected by now, right ?

Nov 12 '05 #167

Noons

Blair Adamache allegedly said,on my timestamp of 10/06/2004 1:15 AM:

And please let everyone refrain from searches like "Blair
Adamache"+wanker or the same for Daniel, Serge or Nuno.

Are you kidding? Been done before. In much more serious places
than this....

--
Cheers
Nuno Souto
wi*******@yahoo.com.au.nospam

Nov 12 '05 #168

Noons

Serge Rielau allegedly said,on my timestamp of 10/06/2004 1:02 AM:

I disagree. Everyone equates Oracle (no qualifier) with the Relational
database, yet the same name is used for Apps, and of course the mobile
database offering.
and quite wrongly, IMHO. Been one of my old grudges with Oracle,
as a matter of fact. It's as stupid as it gets and only confuses people.
Let's not go into what it does to the heads of competition's marketeers.
(the expression "major meltdown" springs to mind...)

The names were originally Oracle Financials for the apps and Oracle RDBMS
for the database. Which would equate to IBM "this" or IBM "that", given
that Oracle is the name of the company. NOT the name of the product.
Somewhere in the last 12 years it all became mixed up, with dire
consequences for intelligible conversation.
The lastest crime is "for Muliplatforms" which sounds like Milla
Jovovich in the "5. Element" and is largely ignored.
Et tu, Brutus? ;)
We had similar problems with "SQL Procedural Langage" (SQL PL) which
isn't a language at all, really.. It's just SQL.
Yeah, that sounded always like a major twist of PL/SQL... :)
I never understood why not just call it what it is: ANSI SQL.
After all, IBM always called Cobol as ANSI Cobol...
Things need short memorisable names, the market demands it, and if there
is none chaos prevails.
It's an IBM blindspot.

Not just IBM...

--
Cheers
Nuno Souto
wi*******@yahoo.com.au.nospam

Nov 12 '05 #169

Mark Townsend wrote:

"SQL" means MS SQL Server

Something I'm sure MS are more than happy with.
UDB means DB2 UDB for Linux Unix and Windows.

This I just don't buy. You say this, but the IBM web site refers to
DB2 Universal Database for .... when referring to all three code bases.

And you have documents entitled "DB2 UDB Family On Common Ground".

And these documents say things like "In February of 1999, IBM® announced
DB2® Universal Database for iSeries (DB2 UDB for iSeries), ..."

Searching for all the following keywords on the IBM website returns vast
numbers of hits
DB2 UDB iSeries - 8,691 hits
DB2 UDB OS/390 - 8,095 hits
Note that the total nbr of hits for DB2 UDB is only 56,445
This is not accidental or even the result of a legacy. I firmly believe
that somebody somewhere at IBM is more than happy that some people read
UDB and mistakenly believe that it's all the same product.

Otherwise all this stuff would have been corrected by now, right ?

Mark, I can't climb into other peoples minds,and I won't even try that
with executives.
But if DB2 UDB is best known on the Linux, Windows and Unix platform and
has that mindshare -
And Oracle is best known for databases, don't you think IBM executives
are as happy as Oracle to implicitly leverage brand-image and carry it over?
As you point out yourself, IBM's message there is inconsistent. It's not
IMHO ignoring the fact that they are different bases _systematically_
which is what is alleged.
I've never encountered having an publication of mine being polished to
exclude the platform identifier. I have seen teh reverse a lot.
If one submitts an article to IDUG or DBM Tech it is _mandatory_ to
clarify the applicability in the title if a talk is not cross-platform.

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

Nov 12 '05 #170

Noons wrote:

Serge Rielau allegedly said,on my timestamp of 10/06/2004 1:02 AM:
I disagree. Everyone equates Oracle (no qualifier) with the Relational
database, yet the same name is used for Apps, and of course the mobile
database offering.

and quite wrongly, IMHO. Been one of my old grudges with Oracle,
as a matter of fact. It's as stupid as it gets and only confuses people.
Let's not go into what it does to the heads of competition's marketeers.
(the expression "major meltdown" springs to mind...)

The names were originally Oracle Financials for the apps and Oracle RDBMS
for the database. Which would equate to IBM "this" or IBM "that", given
that Oracle is the name of the company. NOT the name of the product.
Somewhere in the last 12 years it all became mixed up, with dire
consequences for intelligible conversation.

The rub, I think, is that brands are in companies hands. It has
flipsides, too. E.g. IBM is a "hardware company" DB2 is a "mainframe
product". As much as using DB2 (or Oracle) may help on one side, it
hurts on another.
It is a rather amusing twist that "UDB" is perceived as being DB2 for
LUW. It's a sign of popularity, IMHO. DB2 UDB for LUW has stepped out of
DB2 for z/OS shadow.
The lastest crime is "for Muliplatforms" which sounds like Milla
Jovovich in the "5. Element" and is largely ignored.

Et tu, Brutus? ;)

I tried. Gave up when I figured I'm standing alone and noone nows what
I'm talking about. Back to LUW, it's prettier anyway: DB2 for "love" :-)

We had similar problems with "SQL Procedural Langage" (SQL PL) which
isn't a language at all, really.. It's just SQL.

Yeah, that sounded always like a major twist of PL/SQL... :)
I never understood why not just call it what it is: ANSI SQL.
After all, IBM always called Cobol as ANSI Cobol...

We tried exactly that. But how do tell a customer that:
IBM Rep (eager): "DB2 UDB V7.1 for LUW now supports procedures written
in SQL"
Customer(oblivious): "You mean: "EXEC SQL SELECT c1 INTO var FROM T;"?
What's new about that?"
IBM Rep (excited): "No! No C or Java, just SQL!"
Customer: "But how to I do procedural logic ???"
"In SQL!"
"???*blink*"
Reason: SQL is "perceived" to deal with queries only.
So we settled for SQL "Procedural Language" statements to classify IF,
WHILE etc. and SQL PL is the obvious acronym to use.

Things need short memorisable names, the market demands it, and if
there is none chaos prevails.
It's an IBM blindspot.

Not just IBM...

Possibly, I never worked for another company.

--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

Nov 12 '05 #171

Jeroen van den Broek

"Serge Rielau" <sr*****@ca.eye-be-em.com> schreef in bericht
news:ca**********@hanover.torolab.ibm.com...

Noons wrote:
Serge Rielau allegedly said,on my timestamp of 10/06/2004 1:02 AM:
I disagree. Everyone equates Oracle (no qualifier) with the Relational
database, yet the same name is used for Apps, and of course the mobile
database offering.

and quite wrongly, IMHO. Been one of my old grudges with Oracle,
as a matter of fact. It's as stupid as it gets and only confuses people.
Let's not go into what it does to the heads of competition's marketeers.
(the expression "major meltdown" springs to mind...)

The names were originally Oracle Financials for the apps and Oracle RDBMS for the database. Which would equate to IBM "this" or IBM "that", given
that Oracle is the name of the company. NOT the name of the product.
Somewhere in the last 12 years it all became mixed up, with dire
consequences for intelligible conversation.

The rub, I think, is that brands are in companies hands. It has
flipsides, too. E.g. IBM is a "hardware company" DB2 is a "mainframe
product". As much as using DB2 (or Oracle) may help on one side, it
hurts on another.
It is a rather amusing twist that "UDB" is perceived as being DB2 for
LUW. It's a sign of popularity, IMHO. DB2 UDB for LUW has stepped out of
DB2 for z/OS shadow.

It's not that, I think. It's more because us MF-dinosaurs are tired of
having each and every product renamed all the time; OS/VS, MVS, MVS/XA,
MVS/ESA, OS/390, z/OS and on top of that DB2 on ..., UDB for ..., what's
next? Just give us a name and stick to it.
And the reason for which IBM is still perceived as a "hardware company" is
just the opposite: here they hang on to "International Business Machines",
which doesn't sound like anything but hardware to me...
Maybe they should consider a gradual name-change for the company, something
like "Universal Business Machines" next year, "Universal Data Machines" the
year after, and finally "Universal/Ultimate Data Business"... :-)
After that, you 'truely' can say UDB is UDB is UDB, just like "the
competitor" ;-)

The lastest crime is "for Muliplatforms" which sounds like Milla
Jovovich in the "5. Element" and is largely ignored.

Et tu, Brutus? ;)

I tried. Gave up when I figured I'm standing alone and noone nows what
I'm talking about. Back to LUW, it's prettier anyway: DB2 for "love" :-)
We had similar problems with "SQL Procedural Langage" (SQL PL) which
isn't a language at all, really.. It's just SQL.

Yeah, that sounded always like a major twist of PL/SQL... :)
I never understood why not just call it what it is: ANSI SQL.
After all, IBM always called Cobol as ANSI Cobol...

We tried exactly that. But how do tell a customer that:
IBM Rep (eager): "DB2 UDB V7.1 for LUW now supports procedures written
in SQL"
Customer(oblivious): "You mean: "EXEC SQL SELECT c1 INTO var FROM T;"?
What's new about that?"
IBM Rep (excited): "No! No C or Java, just SQL!"
Customer: "But how to I do procedural logic ???"
"In SQL!"
"???*blink*"
Reason: SQL is "perceived" to deal with queries only.
So we settled for SQL "Procedural Language" statements to classify IF,
WHILE etc. and SQL PL is the obvious acronym to use.
Things need short memorisable names, the market demands it, and if
there is none chaos prevails.
It's an IBM blindspot.

Not just IBM...

Possibly, I never worked for another company.

--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

Nov 12 '05 #172

FWIW, I think we agree violently.

--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

Nov 12 '05 #173

Serge Rielau wrote:

You mean this:
http://www-1.ibm.com/industries/fina...270311103.html

"* Multiple operating systems, running on IBM and non-IBM hardware
* A single code base to deliver a consistent set of functionality
from small single CPU systems, to SMP, to clusters of either"

That statement is correct.

One can pick on the fact that the author didn't fully qualify "DB2 UDB"
to "DB2 UDB for Linux Unix and Windows".'

Cheers
Serge

Lets take this one step at a time since you want to approach this like
an attorney ...

First ... you now acknowledge that IBM did, in fact, write the phrase
"single code base".

Second ... this issue is not the phrase itself, but rather how it is
to be interpreted.

Third ... it is not what was explicitly said ... but rather what was
not said upon which you are making the distinction.

Not every fine slip is malicious? How about every slippery
interpretation? Seems to me the ethical thing to do would be just
to acknowledge that IBM said it and it is pure marketing hyperbole
without basis in fact.

Then we can all get back to postings about things of import such as
Pink Floyd vs. Bach.

--
Daniel Morgan
da******@x.washington.edu
(replace 'x' with a 'u' to reply)

Nov 12 '05 #174

Daniel Morgan wrote:

Then we can all get back to postings about things of import such as
Pink Floyd vs. Bach.

Couldn't we just do this anyway? Don't you all know the second law of
Usenet?

Arguing on the Internet is like running in the Special Olympics. Even if you
win, you are still retarded.

I'm waiting to see who's going to be the first to lose the argument
according to Godwin's Law (first law of Usenet).

Take it offline or meet for a macho-cafe-latte-decaf-muggachino in a
Starbucks somewhere, so you can sort out this vital issue once and for all.

Is c.d.i destined to become a desolate wasteland of pointless arguments and
whingers pining for the good old days?

Nov 12 '05 #175

Jean-David Beyer

Andrew Hamm wrote (in part):

Is c.d.i destined to become a desolate wasteland of pointless arguments and
whingers pining for the good old days?

I am not pining for IBM IMS/DB and CICS, or IBM IMS/DB and IMS/DC. Is
anyone? I did get stuck on a project where we were required to use IBM
IMS/DB and /DC and had to write the censored thing in IBM Basic Assembler
Language ("for efficiency reasons"). They had 70 programmers on that for
about 4 years before the project was scrapped. It may have been efficient,
but since it solved the wrong problem, it was no use to anyone. I argued
in favor of using a higher level language, but was shouted down. With a
higher level language, we could have realized we were solving the wrong
problem a lot sooner.

The good old days! Bah!

--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ Registered Machine 241939.
/( )\ Shrewsbury, New Jersey http://counter.li.org
^^-^^ 22:15:00 up 3 days, 7:40, 8 users, load average: 2.20, 2.19, 2.12

Nov 12 '05 #176

Captain Pedantic

"Andrew Hamm" <ah***@mail.com> wrote in message
news:2i************@uni-berlin.de...

Is c.d.i destined to become a desolate wasteland of pointless arguments and whingers pining for the good old days?

Did you answer your own question there ...?

Nov 12 '05 #177

Daniel,

Let me assert to you that all the code I have written in all my 7 years
for IBM DB2 UDB for Linux, Unix and Windows is single code-base and that
single codebase scales from small single CPU systems (like my laptop),
to SMP (like my development server), to clusters of either (like my
development server cluster of a dozen machines) and across multiple
operating systems (like Linux, Windows, Sun OS, HP, AIX) and hardware
platforms (e.g x (linux, Windows), p (Linux and AIX), z (Linux) and
i(Linux) Series, HP and Sun).
IBM DB2 UDB for Linux, Unix and Windows has a special layer called OSS
which deals with platform specific optimizations and APIs. Depending
whom you ask this isolated layer is between 5%-10% of the codebase.

Given that I touch IBM DB2 UDB for Linux, Unix and Windows code 340 days
a year for 7 years now I happen to know what I am talking about and you
do not.

If you believe that any given company is doing misleading or false
marketing then there are laws in many countries governing this which you
can invoke.

However, there is no doubt that any single phrase one picks out of any
document or source can and will be taken out of context.
It is a base fact in life that this happens and anyone who ever played
"Telephone" ("Stille Post" in my days) knows it's true.
It results in equally smart and objective people comming to widely
different conclusions. Nothing to be concerned about.
And, let's face it, opinions may vary.

I can take a lot of pokes about DB2 and more often than not there are
grains of truth in them (like wouldn't it be nice if all DB2 extenders
were supported on all platforms, as Mark T. likes to rub in ever so gently).

What upsets me about your posts in particluar is that you seemingly
purposely take things out of context and twist them around, you get
corrected over and over and over again and yet you refuse to admit erratas.
And that's where freedom of speech and healthy debate veers of into slander.

What I can't get is, how you don't seem to realize that you not only
loose reputation in non Oracle newsgroups (you may not care), but also
within your "home" Oracle group.
Any of your students, employers, customers and peers can google for
these posts. Don't you care?

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

Nov 12 '05 #178

Madison Pruet

"Daniel Morgan" <da******@x.washington.edu> wrote in message
news:1086913318.264053@yasure...

Then we can all get back to postings about things of import such as
Pink Floyd vs. Bach.
I'm a Mozart fan myself... ;-)

--
Daniel Morgan
da******@x.washington.edu
(replace 'x' with a 'u' to reply)

Nov 12 '05 #179

Serge Rielau wrote:

Daniel,

Let me assert to you that all the code I have written in all my 7 years
for IBM DB2 UDB for Linux, Unix and Windows is single code-base and that
single codebase scales from small single CPU systems (like my laptop),
to SMP (like my development server), to clusters of either (like my
development server cluster of a dozen machines) and across multiple
operating systems (like Linux, Windows, Sun OS, HP, AIX) and hardware
platforms (e.g x (linux, Windows), p (Linux and AIX), z (Linux) and
i(Linux) Series, HP and Sun).
IBM DB2 UDB for Linux, Unix and Windows has a special layer called OSS
which deals with platform specific optimizations and APIs. Depending
whom you ask this isolated layer is between 5%-10% of the codebase.

Serge

And please let me asset to you that code I have written in my 35+ years,
many of which were IBM DB2 on mainframes ... that code could not be
moved, without modification, to a Windows or other platform. Even
basic things like how many characters long is the name of a table
would/could change.

Is that no longer true?

I find it hard to believe that anyone would assert that DB2 on OS/390
where it is shared-everything can be put onto a different operating
system where the DB2 architecture is shared-nothing and think it will
run as is. Please tell me if I am incorrect.

You see the one nice thing about some of DB2's competitors is that
the code written is 100% compatible across operating systems ... not
even requiring a recompilation.

--
Daniel Morgan
http://www.outreach.washington.edu/e...ad/oad_crs.asp
http://www.outreach.washington.edu/e...oa/aoa_crs.asp
da******@x.washington.edu
(replace 'x' with a 'u' to reply)

Nov 12 '05 #180

Mark A

"Daniel Morgan" <da******@x.washington.edu> wrote in message
news:1086927557.861975@yasure...

And please let me asset to you that code I have written in my 35+ years,
many of which were IBM DB2 on mainframes ... that code could not be
moved, without modification, to a Windows or other platform. Even
basic things like how many characters long is the name of a table
would/could change.

Is that no longer true?

I find it hard to believe that anyone would assert that DB2 on OS/390
where it is shared-everything can be put onto a different operating
system where the DB2 architecture is shared-nothing and think it will
run as is. Please tell me if I am incorrect.

You see the one nice thing about some of DB2's competitors is that
the code written is 100% compatible across operating systems ... not
even requiring a recompilation.

--
Daniel Morgan

No, that is not true. The OS/390 (MVS or whatever) COBOL compiler you might
have used with Oracle on OS/390 is not compatible with any Windows, Linux,
or UNIX COBOL compiler (assuming you even wanted to use COBOL on those
platforms).

Initially, there were many incompatibilities between DB2 on OS/2 (the
predecessor of DB2 for Linux, UNIX, and Windows). Those differences at the
DML level are very few now (especially with the release of DB2 version 8 for
z/OS).

The reason that Oracle uses the same code base for all platforms has nothing
to do with portability, it has to do getting a product out the door with a
few changes as possible to work on a particular OS. The fact that Oracle on
does not work very well on OS/390 is a testament to the fact that such a
strategy is mostly a marketing one, and did not produce excellent products
that customers want (at least not on OS/390).

The portability issue is mostly a red herring. Microsoft SQL Server is
selling quite well even though it only runs on Windows. On the platforms
that DB2 competes vigorously with Oracle (Linux, UNIX, and Windows) the DB2
code base is identical on all those platforms.

Nov 12 '05 #181

Mark A wrote:

"Daniel Morgan" <da******@x.washington.edu> wrote in message
news:1086927557.861975@yasure...
And please let me asset to you that code I have written in my 35+ years,
many of which were IBM DB2 on mainframes ... that code could not be
moved, without modification, to a Windows or other platform. Even
basic things like how many characters long is the name of a table
would/could change.

Is that no longer true?

I find it hard to believe that anyone would assert that DB2 on OS/390
where it is shared-everything can be put onto a different operating
system where the DB2 architecture is shared-nothing and think it will
run as is. Please tell me if I am incorrect.

You see the one nice thing about some of DB2's competitors is that
the code written is 100% compatible across operating systems ... not
even requiring a recompilation.

--
Daniel Morgan

No, that is not true. The OS/390 (MVS or whatever) COBOL compiler you might
have used with Oracle on OS/390 is not compatible with any Windows, Linux,
or UNIX COBOL compiler (assuming you even wanted to use COBOL on those
platforms).

Initially, there were many incompatibilities between DB2 on OS/2 (the
predecessor of DB2 for Linux, UNIX, and Windows). Those differences at the
DML level are very few now (especially with the release of DB2 version 8 for
z/OS).

The reason that Oracle uses the same code base for all platforms has nothing
to do with portability, it has to do getting a product out the door with a
few changes as possible to work on a particular OS. The fact that Oracle on
does not work very well on OS/390 is a testament to the fact that such a
strategy is mostly a marketing one, and did not produce excellent products
that customers want (at least not on OS/390).

The portability issue is mostly a red herring. Microsoft SQL Server is
selling quite well even though it only runs on Windows. On the platforms
that DB2 competes vigorously with Oracle (Linux, UNIX, and Windows) the DB2
code base is identical on all those platforms.

Thanks for the clarification.

The reason Oracle doesn't concentrate on OS/390, if that is the case,
might also relate to the difficulty of selling into a marketshare very
carefully managed by IBM. I know that was the case when I last worked
on that class of machine.

--
Daniel Morgan
http://www.outreach.washington.edu/e...ad/oad_crs.asp
http://www.outreach.washington.edu/e...oa/aoa_crs.asp
da******@x.washington.edu
(replace 'x' with a 'u' to reply)

Nov 12 '05 #182

Mark A

"Daniel Morgan" <da******@x.washington.edu> wrote in message
news:1086929808.312585@yasure...

Thanks for the clarification.

The reason Oracle doesn't concentrate on OS/390, if that is the case,
might also relate to the difficulty of selling into a marketshare very
carefully managed by IBM. I know that was the case when I last worked
on that class of machine.

--
Daniel Morgan

You make it sound like IBM has the power of the mafia. I don't think so.

Actually, a surprising number of companies have purchased Oracle for OS/390.
Oracle gave me 5 references in the Fortune top 25 companies in the US. I
checked all 5 out, and all 5 purchased it, but not one was using it for a
production system.

Nov 12 '05 #183

If one submitts an article to IDUG or DBM Tech it is _mandatory_ to
clarify the applicability in the title if a talk is not cross-platform.

Here are session titles from this years IDUG Conferences - very few of
them cross platform, and not a single one with any sign of platform
applicability. I guess we need to differ of what _mandatory_ means as well.

2PC: DB2 is from Mars & WebServers are from Venus
60 DB2 UDB Tips in 60 Minutes
A List of DB2 Top Ten Lists In which we ponder numerouse DB2 topics for
learning
Amaze Your Friends and Impress Your Family: More Details of DB2 UDB
Internals
An insider's look at the next version of DB2
And I Thought I Knew all There Was to Know about DB2
Archiving Data into an “Offline” DB2 Database
Are your DB2 Tools ready for the latest version of DB2 ?
Automated DB2 Release Migrations
Automated Statistics Collection with DB2 V8.2
Automating DB2 Structure Changes
Autonomic DB2-simplicity/performance, availability
Availability 101 for DB2
Benchmarking a DB2 Application
Best practices for managing DB2 Large Objects
Building and Maintaining DB2 UDB Data Warehouses
Considering "Dipsies" - New Indexes for DB2 V8
Continuous Data Protection in DB2 Environments
DB2 - Ready for Java
DB2 and the Internet – 17 Performance Tips
DB2 And Workload Manager: Setting Optimal Goals
DB2 and XML
DB2 Application Development - v8 update
DB2 Application Development; Triggers / Stored Procedures / UDF’s
DB2 Application Development; XML / .NET / JAVA / Web Services
DB2 B&R Speed is still the most critical demand
DB2 Catalog and Migration: past, present & future
DB2 Cube Views: the OLAP-aware RDBMS
DB2 Design Advisor: More than just index selection
DB2 HA with VERITAS Cluster Products
DB2 in a .NET World
DB2 Kernel - The Inside Tour
DB2 Memory: Impacts on Application Performance.
DB2 Stored Procedures Masterclass
DB2 UDB and Grid Computing
DB2 UDB High Performance Design
DB2 UDB in 2004. A Technology Update!
DB2 UDB Internals : Indepth (In 2-days)
DB2 UDB Internals for Administrators : Complete with V8 Details
DB2 UDB V8 : Exploiting it's New Advancements in High Availability
DB2 UDB V8 Database Administration Certification Preparation
DB2 V8 SQL and Performance Enhancements
DB2 V8 Stored Procedures and WLM – Get Ready Now!
DB2 V8 System Point-in-Time Recovery for ERP
DB2 Version 8 - Partitioning Changes
DB2 Version 8 for Application Developers
Designing and building a 24X7 database using DB2
End-to-End DB2 UDB Performance and Management Solutions
Enterprise DB2 Monitoring On The Cheap
Fast Track To Optimal DB2 Performance
Get Ready for DB2 V8 – Improving Access Paths without Risk
Get the Latest on DB2 V8 Straight From the Source – IBM’s DB2 Lead
Strategist
Getting More Kicks out of DB2 Performance Data
Getting the Most out of UDB Monitoring and Tuning
Hands-on programming - DB2 Application Development
Health Checking Your DB2 System
High Availability and Disaster Recovery in DB2 UDB
HOW DB2 HAS FACED A DISASTER RECOVERY SIMULATION
How To Avoid Common DB2 Implementation Mistakes
How to compare instances of DB2 Objects.
IBM DB2 Administration and Performance Tools
IBM DB2 Cube Views -- Large Customer Case Study
Implementing and Monitoring DB2 UDB replication
Inside DB2's Backup and Recovery
Integrating EMC Storage Technologies into DB2 UDB Database Solutions
JDBC Best Practices for DB2 Programmers
Key design in DB2
Managing Query Workload with your DB2 UDB Database
Memory Management in DB2 UDB - V8 Updates
Monitoring and Tracking Availability with UDB
OLAP in DB2: MDCs, MQTs, and Cube Views
Performance Tuning in Large DB2 Environments
Practical Techniques for DB2 Application Tuning
Recovery Management for DB2
SAP and DB2 : Mission Impossible or Pretty Woman ?
SQL in DB2 UDB V8: Changes, Challenges and Quirks
Test Data Privacy and DB2 Application Testing Solutions
The Doctor is In! Advanced Performance Diagnostics in DB2
The Frugal Data Warehouse - Doing more with DB2
The Ties That BIND DB2
Time-Delayed Log-Based DB2 UDB Replication with Aivant DBShadow
Too Much Data in my DB2! What is Active Archiving?
Tuning DB2 for the New Galaxy
Tuning your DB2 Pools How much memory do you Really Need?
Unicode and DB2 V8, what you need to know
Ways to access DB2 from a Java/web application
What's new (and coming) in DB2 UDB in 2004.

Nov 12 '05 #184

Jean-David Beyer wrote:

I am not pining for IBM IMS/DB and CICS, or IBM IMS/DB and IMS/DC.
Is anyone?

hmmm - i also forgot to allow for the cross-posting. I'm talking about the
steady stream of nay-sayers that are still having panic attacks after the
IBM takeover of Informix.

Although I agree that IBM is a major player in the IT Axis of Evil, I also
believe that IBM is the bestest thing that could have happened to Informix,
apart, perhaps, from Informix itself keeping it's shit together over the
last few years instead of being abused by corporate raiders etc.

Some people are still remembering the Alamo a little too much.

Nov 12 '05 #185

Captain Pedantic wrote:

"Andrew Hamm" <ah***@mail.com> wrote in message
news:2i************@uni-berlin.de...
Is c.d.i destined to become a desolate wasteland of pointless
arguments and whingers pining for the good old days?

Did you answer your own question there ...?

Never let irony stand in the way of a good story.

Nov 12 '05 #186

Actually, a surprising number of companies have purchased Oracle for OS/390.
Oracle gave me 5 references in the Fortune top 25 companies in the US. I
checked all 5 out, and all 5 purchased it, but not one was using it for a
production system.

Oracle for OS/390 was more widely used 5-10 years ago when we saw a lot
of migration of applications from mainframes to open system platforms -
typically deployed as an aid to migration supporting parallel
testing/development etc. They tended to get "switched off" once the
migration was complete.

Some OS/390 customers also used Oracle on the mainframe as an "in
memory" gateway to their DB2 data for their downstream systems etc. In
some situations the performance of Oracle (Unix) <-> Oracle (OS/390) <->
TG4DB2 (OS/390) <-> DB2 could be better configured than Oracle (Unix)
<-> TG4DB2 (OS/390) <-> DB2 (OS/390) (better control of the join
shipping, for one).

So it's entirely possible that companies have simply stopped using
Oracle OS/390, or only use it as an aid to striping data from the MF
environment. Let me know who the references were and I'll make sure we
stop giving them out.

Nov 12 '05 #187

> The OS/390 (MVS or whatever) COBOL compiler you might

have used with Oracle on OS/390 is not compatible with any Windows, Linux,
or UNIX COBOL compiler (assuming you even wanted to use COBOL on those
platforms).

I seem to remember (from faded memory) that Microfocus had a compiler
that was (reasonably) compatible. Or was that just to the OS/2 ? I know
we moved a swag of Cobol FinServs apps onto HP a decade ago without
significant rewrite. Is microfocs even around anymore - CA-Microfocus ?

Nov 12 '05 #188

Mark A

> > The OS/390 (MVS or whatever) COBOL compiler you might

have used with Oracle on OS/390 is not compatible with any Windows, Linux, or UNIX COBOL compiler (assuming you even wanted to use COBOL on those
platforms).

"Mark Townsend" <ma***********@comcast.net> wrote in message
news:HIbyc.26081$HG.7386@attbi_s53... I seem to remember (from faded memory) that Microfocus had a compiler
that was (reasonably) compatible. Or was that just to the OS/2 ? I know
we moved a swag of Cobol FinServs apps onto HP a decade ago without
significant rewrite. Is microfocs even around anymore - CA-Microfocus ?

Yes MicroFocus is still around, and it was reasonably compatible. But Danny
boy only allows 100% compatibility for DB2, so applying the same standard,
MicroFocus was not 100% compatible with mainframe COBOL.

Nov 12 '05 #189

Noons

Mark A allegedly said,on my timestamp of 11/06/2004 3:02 PM:

You make it sound like IBM has the power of the mafia. I don't think so.
I very much think so, having been on the receiving end of their
sales techniques for a looooooong time....

Actually, a surprising number of companies have purchased Oracle for OS/390.
Oracle gave me 5 references in the Fortune top 25 companies in the US. I
checked all 5 out, and all 5 purchased it, but not one was using it for a
production system.

That to me spells one thing and one thing only: there is a surprising
number of companies that are looking at ANY way of becoming LESS
dependent on IBM. I wonder why...
Unfortunately, there are very few people left around that know
exactly how to take advantage of that.
--
Cheers
Nuno Souto
wi*******@yahoo.com.au.nospam

Nov 12 '05 #190

Noons

Madison Pruet allegedly said,on my timestamp of 11/06/2004 1:09 PM:

Then we can all get back to postings about things of import such as
Pink Floyd vs. Bach.

I'm a Mozart fan myself... ;-)

and brother Ray Charles passed away today.
A minute of silence, boys and girls.
The world is now a smaller place, as far as I'm concerned...
:(

--
Cheers
Nuno Souto
wi*******@yahoo.com.au.nospam

Nov 12 '05 #191

Mark A

"Noons" <wi*******@yahoo.com.au.nospam> wrote in message
news:40**********************@news.optusnet.com.au ...

Mark A allegedly said,on my timestamp of 11/06/2004 3:02 PM:
You make it sound like IBM has the power of the mafia. I don't think so.
I very much think so, having been on the receiving end of their
sales techniques for a looooooong time....

Sales techniques? Maybe you could explain your libelous statement in detail.
I doubt very much that your were in a decision making postion anyway, which
is what you are really complaining about.
Actually, a surprising number of companies have purchased Oracle for OS/390. Oracle gave me 5 references in the Fortune top 25 companies in the US. I
checked all 5 out, and all 5 purchased it, but not one was using it for a production system.

That to me spells one thing and one thing only: there is a surprising
number of companies that are looking at ANY way of becoming LESS
dependent on IBM. I wonder why...
Unfortunately, there are very few people left around that know
exactly how to take advantage of that.
--
Cheers
Nuno Souto

Not exactly. You have a limited imagination. They bought Oracle on a smaller
platform and believed the hype that Oracle could successfully port their
application to any platform. As one consultant told me, the reason that
people buy Oracle on OS/390 (even though it doesn't work well on that
platform) is the same reason that people buy pet rocks. Now that is real
marketing.

Nov 12 '05 #192

Anton Versteeg

Noons wrote:

Mark A allegedly said,on my timestamp of 11/06/2004 3:02 PM:
You make it sound like IBM has the power of the mafia. I don't think so.

I very much think so, having been on the receiving end of their
sales techniques for a looooooong time....

Actually, a surprising number of companies have purchased Oracle for
OS/390.
Oracle gave me 5 references in the Fortune top 25 companies in the US. I
checked all 5 out, and all 5 purchased it, but not one was using it for a
production system.

That to me spells one thing and one thing only: there is a surprising
number of companies that are looking at ANY way of becoming LESS
dependent on IBM. I wonder why...
Unfortunately, there are very few people left around that know
exactly how to take advantage of that.

You confuse two things:
Buying/using a company's product and becoming dependent on a company.
You are right though that customers don't like to be dependent of a
company whether it is IBM, Oracle or MS. That is probably why Linux is
growing rapidly in spite of MS accusations that it is more expensive to
administer. Might also explain why Oracle is growing much less than its
rivals. Ever tried to port SQLForm applications?
--
Anton Versteeg
IBM Certified DB2 Specialist
IBM Netherlands

Nov 12 '05 #193