No future for DB2

[Serge Rielau]

Re: troll

Jurgen Haan wrote:[color=blue]
> Serge Rielau wrote:
>[color=green]
>> Call it RAC, GRID, autonomous computing, queuing, advanced XML,
>> information integration. I think the commercial database vendors will
>> stay ahead of the open source crowd for the foreseeable future.
>> Comparing marketshare between the faction in this environment is
>> another can of worms.
>>[/color]
>
> I DO think you're underestimating both growth rate and capabilities of
> the OS databases. I honestly think those databases will gain a lot of
> marketshare in a very short time.[/color]
<snip>
Let me qualify what I posted.
I think the commercial databse vendors will _technologicall y_ stay ahead
of the open source crowd.
Measuring Marketshare is a very difficuly thing when you add free stuff
to the mix.
What is the market share of ACCESS? What is the marketshare of all those
build-in niche products?
If we go back to the "# of books at Indigo" example ACCESS may well be
market leader.
For all I know the big three are already in the minority in number of
installations/amount of data stored/???. Who knows?

Let me spin it a different way:
What is the "marketshar e" of pedestrians in Toronto as part of the
public and private transportation? 0! Yet the sidewalks are choked.

The question for commercial vendors is: Does the market share as defined
in revenue shrink or grow. So far it has been growing. And that is what
is counting for teh commercial vendors.
As long as the pie for commercial vendors grows there is no threat.

Just my two cents not being an MBA or such...
Serge
--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

[Jurgen Haan]

Re: troll

Serge Rielau wrote:

[color=blue]
> The question for commercial vendors is: Does the market share as defined
> in revenue shrink or grow. So far it has been growing. And that is what
> is counting for teh commercial vendors.
> As long as the pie for commercial vendors grows there is no threat.
>[/color]

By marketshare, is this the share divided over commercial products?
Because OS products cannot be calculated in this pie. no one can measure
the number of OS products in use, number of downloads is no good,
because the OS databases are mirrored and distributed among different
Linux/Unix flavors. To illustrate, no one on the face of this earth
(except the ones who work here) could know we use postgreSQL and Mysql
in our company, since the medium where it's installed from is a copied
slackware distro, and the databases are installed on several servers.
But for registered figures, we only use DB2, because we are a registered
customer with db2.
If you mean that the total amount of customers is growing, then you're
right, that's a good thing.
Considering tech aspect, most OS products tend to surpass their
commercial counterparts quite fast. Only take a look at desktops. I know
it's a dark unknown area and you need a good amount of knowledge and
time to customise a X desktop, but atm I have a desktop on my Laptop
which outperforms (and at least out-eycandies windows) both windows and
MacOS-X and which uses techniques which are to be introduced in the
upcoming Longhorn/Vista. But most people just don't know about it, or
think it's just a collection of toys or beta libs, there's no marketing
behind those products, but still they exist. Take a look at the progress
on the linux kernel, quite impressive. Other examples; Openoffice,
Mozilla, Apache, Samba (yes... samba 3 is technological superiour to the
SMB implem in both win 2003 & XP), where windows CE was a nice mobile
solution, more and more companies are using linux for embedded and
mobile solutions, DVD players, sattelite decoders, pda's, mobile phones,
navigational systems. But nobody knows... No one tells the public: we're
using OS software in our products.
Why do I take those Microsoft products as an example? Microsoft is a
major competitor in the general computer market, they have an enormous
amount of marketshare, but that does not mean their products are good.
Back ontopic: anyone noticed MonetDB? A new (registered @ sf.net in
2002) OS DB product built by the Dutch CWI institute. It uses some new
storage techniques and a transparent interface which enables the
database to use any type of language to access data. Both SQL and XQuery
are default supported. The database is lightspeed compared to both Mysql
and PostgresDB and already it has an UnixODBC interface, and new
features are added every day.

-R-

[Captain Pedantic]

Re: No future for DB2 - slightly off-topic, discusses what people are being taught at uni

"Jurgen Haan" <jurgen@fake.do m> wrote in message
news:42f1dc46$0 $11064$e4fe514c @news.xs4all.nl ...
[color=blue]
> When I was in school, which was a very long time (took me 7 years to
> complete a 4 year education of Advanced Informatics (which has little to
> do with witts, but more with the appealing student life ;)), I saw a very
> disturbing flow. When I started we were educated in Oracle, Unix
> (Solaris/Linux), Overall databases (isolation levels, ACID properties,[/color]

The type of acid properties *I* studied at university didn't have much to do
with databases!

[Serge Rielau]

OS DBMS vs commercial software Was: troll

I took the liberty to adjust the subject to the topic

Jurgen Haan wrote:[color=blue]
> Serge Rielau wrote:
>
>[color=green]
>> The question for commercial vendors is: Does the market share as
>> defined in revenue shrink or grow. So far it has been growing. And
>> that is what is counting for teh commercial vendors.
>> As long as the pie for commercial vendors grows there is no threat.
>>[/color]
>
> By marketshare, is this the share divided over commercial products?[/color]
Correct. That's how it's commonly counted.[color=blue]
> Because OS products cannot be calculated in this pie. no one can measure
> the number of OS products in use, number of downloads is no good,
> because the OS databases are mirrored and distributed among different
> Linux/Unix flavors. To illustrate, no one on the face of this earth
> (except the ones who work here) could know we use postgreSQL and Mysql
> in our company, since the medium where it's installed from is a copied
> slackware distro, and the databases are installed on several servers.
> But for registered figures, we only use DB2, because we are a registered
> customer with db2.[/color]
Right. To gartner you are true blue and as long as your IBM rep doesn't
feel mySQL is eating into space that should better be owned by DB2 IBM
doesn't care.[color=blue]
> If you mean that the total amount of customers is growing, then you're
> right, that's a good thing.[/color]
Absolutely, and as long as the total amount of customesr grows for both
commercial and OS then that's good enough.[color=blue]
> Considering tech aspect, most OS products tend to surpass their
> commercial counterparts quite fast.[/color]
<snip>
I think we largely agree. Not sure I agree on the technological surpassing.
I do not see Linux desktops (yet) outside of the geeky area and I don't
see how Linux has knocked teh commercial OS vendors over technologically .
Note however how Microsoft is struggling to incite customers to upgrade.
There is little to be gained technologically by going from one version
of Windows to the next for the vast majority of desktop users.
To most Windows 2000 was good enough. The moment Linux desktops
environment catch up with W2k and that is widely known I'd expect MS
#installations to tank unless they go free themselves because MS has no
where to go with the desktop.
If you look at DBMS again vendors are trying to avoid the trap by moving
upstream into "Informatio n Integration" and "Content management" (see
IBM's shopping spree) and Apps (Peoplesoft, Retek). Not moving upstream
to provide value means getting clobbered by the mySQL et al.

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

[Jurgen Haan]

Re: No future for DB2 - slightly off-topic, discusses what peopleare being taught at uni

Captain Pedantic wrote:
[color=blue]
>
> The type of acid properties *I* studied at university didn't have much to do
> with databases!
>
>[/color]

*LOL* perhaps for the cmos battery in the DB server :)

-R-

[Jurgen Haan]

Re: OS DBMS vs commercial software Was: troll

Serge Rielau wrote:
[color=blue]
> I do not see Linux desktops (yet) outside of the geeky area ...[/color]
Unfortunately that is true, it's very pleasant to work with linux
desktops, I use it al the time at work and even in private.
But no marketing means (nearly) no customers.
[color=blue]
> ... and I don't
> see how Linux has knocked teh commercial OS vendors over technologically .[/color]
There a number of implementations in the linux OS which are yet to be
announced in future versions of other OS's. Furthermore some
implementations are better handled in The OSS than in the commercial
counterparts. Again I use the example of Samba. Samba has all the
abilities of the microsoft SMB implementation, and then some. Next to
that Samba outperformed win 2003 SMB with a factor 3 in speed. (this is
just one example)
[color=blue]
> Note however how Microsoft is struggling to incite customers to upgrade.
> There is little to be gained technologically by going from one version
> of Windows to the next for the vast majority of desktop users.[/color]
Yes.. this is again obvious in the upcoming Windows Longhorn/Vista.
Compared to the first announcement, a numerous of new revolutionary
parts have been dropped (one of them was bash-like scripting). If you
subtract that from the initial presentation, you'll get a polished XP.
[color=blue]
> To most Windows 2000 was good enough. The moment Linux desktops
> environment catch up with W2k and that is widely known I'd expect MS
> #installations to tank unless they go free themselves because MS has no
> where to go with the desktop.[/color]
This will most likely never happen, at least not through linux.
Almost no one I know has even heard of linux (or they have and don't
even know how to pronounce it, or even call it lunix (c64 unix))
Without marketing, linux will stay what it is; a geeky os.
[color=blue]
> If you look at DBMS again vendors are trying to avoid the trap by moving
> upstream into "Informatio n Integration" and "Content management" (see
> IBM's shopping spree) and Apps (Peoplesoft, Retek). Not moving upstream
> to provide value means getting clobbered by the mySQL et al.[/color]
It would be ludicrous to use a vast system like DB2 or Oracle for use of
a CMS which displays some newsitems. Talk about a waste of potential.
The fact is that the OS db's are shifting towards OLTP (this is the hard
one) and Data warehousing. (another fact is that they are not there yet)
[color=blue]
> Cheers
> Serge[/color]

-R-

[Madison Pruet]

Re: No future for DB2


"Noons" <wizofoz2k@yaho o.com.au> wrote in message
news:42f1ede1$0 $11939$5a62ac22 @per-qv1-newsreader-01.iinet.net.au ...[color=blue]
> Madison Pruet apparently said,on my timestamp of 4/08/2005 9:21 AM:[color=green]
> >
> > Fair enough --- on the IDS informix server, to do this with say 5[/color][/color]
instances[color=blue][color=green]
> > by issueing the two commands
> >
> > cdr define template noons --database=noons --master=node1 --all
> > cdr realize template noons --syncdatasource= node1 node2 node3 node4[/color][/color]
node5[color=blue][color=green]
> >
> > This would set up an update anywhere of everything in the noons database
> > using the five servers.[/color]
>
> Thank you. Like I suspected, you have to relicate the entire database.
> It is much more efficient to do that in Oracle via dataguard. Like
> I said nearly 24 hours ago. Of course, you can also template
> the entire database and replicate the lot. Anything is possible,
> provided you have sufficient resources.[/color]

No Noons. You asked how I would do it, not how it had to be done.

Since the IBM Informix Enterprise Replication uses less than 10% of the
total resources, it is simply easier to replicate the whole DB. And since
disks are so cheap now, I would have difficulty not justifying having a
multi-node grid like arrangement so that the applications could be active on
any of the replicated nodes.

[color=blue]
>
> --
> Nuno Souto
> in sunny Sydney, Australia
> wizofoz2k@yahoo .com.au.nospam[/color]

[rkusenet@yahoo.com]

Re: No future for DB2

"Madison Pruet" <mpruet@comcast .net> wrote
[color=blue]
> Since the IBM Informix Enterprise Replication uses less than 10% of the
> total resources, it is simply easier to replicate the whole DB. And since
> disks are so cheap now, I would have difficulty not justifying having a
> multi-node grid like arrangement so that the applications could be active on
> any of the replicated nodes.[/color]

so what you are saying is that with ER on a multi node replication (all
updating
each other) one can develop a grid aware application which can not only
provide high availability, but also load balancing.

A friend of mine in USA is working on a project like this only. He says
it is incredible.

[Frank van Bortel]

Re: No future for DB2

Noons wrote:[color=blue]
>
> Back in 91 I drove out of Sweden one early morning. Ended up in
> France in the afternoon of the same day. After zooming past
> Denmark, Germany, Holland and Belgium. So I decided it was
> probably worth going back and having a better look at the ones
> I missed when I blinked. Spent the night in Holland. Nice
> pub, near one of the freeways. Owner asked me where I had been
> that day. When I explained he laughed and called me mad. Which
> was probably quite true. Heck, I was in holidays! ;)
>
> (of course over here we often drive 600Ks each way for the
> weekend, but that's us geographically challenged people...)
> <g,d&r>
>[/color]
You are a cheat! Denmark as well as France have common borders
with Germany. And there's no speed limit in Germany.

But yes - it's feasible, no problem. BTW you forgot Luxembourg.
Did 300Ks daily for almost 5 years.
--
Regards,
Frank van Bortel

[Noons]

Re: No future for DB2

Frank van Bortel wrote:[color=blue][color=green]
> >[/color]
> You are a cheat! Denmark as well as France have common borders
> with Germany. And there's no speed limit in Germany.[/color]

Hey, don't blame me: I didn't make the borders! :)
Yeah, German autobahns ROCK! Although I must admit Northern
Italy was the most unsettling: doing 170 from Venice to
Florence when some guy in a "whatever-zoometti" just went
woosh past us: felt like opening the door and getting out!
Tried to catch him but got the foot off the pedal at 240.
It was a holiday after all and I wanted to enjoy it till
the end... (darn 2L Sierras could MOVE!)
[color=blue]
> But yes - it's feasible, no problem. BTW you forgot Luxembourg.[/color]

Missed it. I think I caught a glimpse on the rear-mirror.
;)
[color=blue]
> Did 300Ks daily for almost 5 years.[/color]

You're MAD! :)
Wouldn't try that even here. Well maybe in the NT or SA
where there are no speed limits on outback roads...

[Noons]

Re: No future for DB2

Madison Pruet wrote:[color=blue]
> Since the IBM Informix Enterprise Replication uses less than 10% of the
> total resources, it is simply easier to replicate the whole DB. And since
> disks are so cheap now, I would have difficulty not justifying having a
> multi-node grid like arrangement so that the applications could be active on
> any of the replicated nodes.[/color]

Pretty soon you're gonna try to convince me replication
on IIER is faster than light...

[bka]

Re: No future for DB2

IBM, the company that invented SQL and relational database, is tied
(according to Gartner) for relational database revenue with Oracle. The
article that spawned this thread is a cheapshot by a nonentity seeking
self-promotion. I can't say much about Mac, but the characteristics
identified by Charles Darwin for the survival of a species are inherent
in DB2.

[Superboer]

Re: No future for DB2

the article may have some truth in it; may be it's a warning
that IBM is suffering from the Philips syndrome; producing
great technology but may loose the market.

If that is the case then it is a very good idea to take these signals
serious and start doing something about it.

Superboer.



bka schreef:
[color=blue]
> IBM, the company that invented SQL and relational database, is tied
> (according to Gartner) for relational database revenue with Oracle. The
> article that spawned this thread is a cheapshot by a nonentity seeking
> self-promotion. I can't say much about Mac, but the characteristics
> identified by Charles Darwin for the survival of a species are inherent
> in DB2.[/color]

[Larry]

Re: No future for DB2

I duuno ... don't see any signs that IBM is losing significant market
share. Look at other examples in various industries ... for many
products, there are usually at least 2 or 3 or 4 main survivors who end
up there for the long-term. You don't even have to be the market share
leader ... just have a significant presence. Hertz/Avis,
American/United/Delta, Krups/Braun/Cuisinart, pick your example.

And, in fact, IBM's product and portfolio diversification as well as
research funding should be seen as advantages to long-term survival over
companies with less diverse product lines and solution lines.

Larry

Superboer wrote:[color=blue]
> the article may have some truth in it; may be it's a warning
> that IBM is suffering from the Philips syndrome; producing
> great technology but may loose the market.
>
> If that is the case then it is a very good idea to take these signals
> serious and start doing something about it.
>
> Superboer.
>
>
>
> bka schreef:
>
>[color=green]
>>IBM, the company that invented SQL and relational database, is tied
>>(according to Gartner) for relational database revenue with Oracle. The
>>article that spawned this thread is a cheapshot by a nonentity seeking
>>self-promotion. I can't say much about Mac, but the characteristics
>>identified by Charles Darwin for the survival of a species are inherent
>>in DB2.[/color]
>
>[/color]

[Chris \( Val \)]

Re: No future for DB2


"Noons" <wizofoz2k@yaho o.com.au> wrote in message
news:42ec2db5$0 $11917$5a62ac22 @per-qv1-newsreader-01.iinet.net.au ...
| DA Morgan apparently said,on my timestamp of 31/07/2005 3:32 AM:
|
| replying to both:

[snip]

| Maybe we'll see some management finally biting the bullet and
| cleansweeping the Java crap off the place? Sure, it'll have to
| be washed with liberal applications of marketing bullshit.
| That shouldn't be a problem.

:-)

http://www.dack.com/web/bullshit.html

Cheers,
Chris Val