Stefaan A Eeckels <tengo@DELETEMEecc.lu> wrote:
> On Wed, 8 Jun 2005 10:30:09 +0000 (UTC)
> Przemek Bak <przemolicc@poczta.fm.-no-spam.invalid> wrote:
>
> > How about Sybase ?
>
> Why buy a database company? The problem with DBMSes is the
> same as with Operating Systems: unless your the dominant
> player, charging money for them is very difficult.

But Sun could have better portfolio and could better integrate
everything.
Do you want to have big RDBMS ?
Go to Sun.
Do you want to have big server ?
Go to Sun.
Do you want to have big storage ?
Go to Sun.
Do you want to have big RDBMS on big server with big storage ?
Go to Sun.

przemol

On 2005-06-08, John D Groenveld <groenvel@cse.psu.edu> wrote:
> Can't wait to see Sun's mobile workstation system.

Once upon a time, there was a SPARCstation Voyager [1].

Andreas.

[1] http://www.bsn.com/BSN/Support/FAQS/Voyager.html

On Thu, 9 Jun 2005 21:39:38 +0000 (UTC)
groenvel@cse.psu.edu (John D Groenveld) wrote:

> In article <20050609223241.769490c2.tengo@DELETEMEecc.lu>,
> Stefaan A Eeckels <tengo@DELETEMEecc.lu> wrote:
> >Are you suggesting that Sun should have been more like IBM
> >proposing its own integrated service and software? I can
>
> Yes.
>
> >understand that going head-to-head with Oracle (which sold
> >a lot of systems for Sun) during the boom would have been
> >considered foolish (especially by Wall Street pundits).
>
> Why isn't IBM considered foolish by pundits for competing with
> Oracle by selling DB2 on IBM, Sun, and other company's systems?

Maybe because they developed DB2 instead of buying an ailing
DBMS outfit at the time Oracle was building up its business?

> Why does Oracle sell to IBM system customers?

Because they like to earn money? I'm sure that Oracle would
not refuse to sell to Sun customers were Sun to have AES in
its portfolio. That being said, they might have been less
enthusiastic in their support of Sun if Sun had started to
compete with them in the late 1990ies.

> >Or are you suggesting that Sun should have been a database
> >company? Well, the guys who started it could build computers.
>
> No. I'm suggesting Sun should stick to being a systems company that
> also sells hardware as well as an expanded line of software.

Which they do. Only they do not sell a database. It's easy
to say with 20/20 hindsight they should have had a database
since at least 2000, but remember that by that time Oracle
had largely won the battle, so buying any of the remaining
losers might not have revived its chances.

The fact that Sun did not develop its own database is because
they started as a workstation company. I can understand them
not buying one of the also-rans in the second half of the
1990ies. I still cannot see any advantage to buying Sybase now.

> >boxes even run Windows). To continue to sell in the high-end
> >market, SPARC systems will need to get as close as possible
> >in speed to x86 (and there's hope there, as Intel has indicated
> >that there is a limit to clock speed), or Opteron will have to
> >scale to SPARC levels.
>
> How do you define speed and how does your definition compare with
> Sun's definition of throughput?
> <URL:http://www.google.fi/search?q=throughput+computing+site%3Ablogs.sun.com &hl=en>

The fact of the matter is that many of the tasks you did need
a large Sun for in the 1990ies can now be performed by cheap
x86 boxes running Linux. If they do not keep the lower-end
SPARC servers competitive with the Dells of this world, people
will migrate to x86 hardware, and not necessarily Suns' Opteron
boxes. After all, you also need to recompile to move from
Solaris/SPARC to Solaris/x86.

Whether Sun is right with its appreciation of throughput or not
doesn't matter because IT managers buy their machines based on
either artificial benchmarks (the rational ones) or pure gut
feel. I know of a large organisation that uses lots of large
and medium SPARC servers running Solaris, but will not consider
Solaris/x86 for the ever increasing horde of low-end x86 servers,
where they mandate Linux. Go figure.
>
> >How is buying database companies with less than 3% market share
> >going to suddenly become a major source of revenue for Sun?
>
> Why would have Sun ASE been doomed to only have 3% of the database
> market?

That's something no-one can answer. It's better than Oracle in
many ways, but since MS-DOS and Windows we know that technical
excellence means diddly squat.

> How is it a good thing for Sun that Sun relies on Oracle and IBM to
> provide the database functionality that is the core of many
> applications?
>
> Has IBM and Oracle's pricing been good for Sun customers?

Oracle's pricing is designed to extract as much money as possible
from customers, meaning that if they spend less on hardware, they
can spend more on software. This is not good for hardware vendors.

I don't follow IBM's DB2 pricing on Solaris. Do they lower the
price of DB/2 if people by a P-series?

Still, this does not mean that Sun offering AES for a lower fee
would necessarily stop people from buying Oracle, or DB/2.

> >People who are cheap can get free stuff already.
>
> Oracle, even on clusters of single core hardware, isn't free.

Which is why people tend to buy the cheapest machines they
perceive can do the job of running Oracle. Those who do not
want to spend money already have alternatives (MySQL for
example) and mostly, they would also tend to buy the cheapest
machines they can get away with.

The problem is that machines have become so powerful that most
of the things can be done by cheap x86 boxes running Linux.

--
Stefaan
--
As complexity rises, precise statements lose meaning,
and meaningful statements lose precision. -- Lotfi Zadeh

On Fri, 10 Jun 2005 08:15:15 +0000 (UTC)
Przemek Bak <przemolicc@poczta.fm.-no-spam.invalid> wrote:

> But Sun could have better portfolio and could better integrate
> everything.
> Do you want to have big RDBMS ?
> Go to Sun.
> Do you want to have big server ?
> Go to Sun.
> Do you want to have big storage ?
> Go to Sun.
> Do you want to have big RDBMS on big server with big storage ?
> Go to Sun.

Indeed. Unfortunately, AES is not widely perceived as
"a big RDBMS", and Sun cannot afford Oracle.

--
Stefaan
--
As complexity rises, precise statements lose meaning,
and meaningful statements lose precision. -- Lotfi Zadeh

In article <20050610141535.770baeb3.tengo@DELETEMEecc.lu>,
Stefaan A Eeckels <tengo@DELETEMEecc.lu> wrote:
>Maybe because they developed DB2 instead of buying an ailing
>DBMS outfit at the time Oracle was building up its business?

Why did IBM buy Informix?

>Because they like to earn money? I'm sure that Oracle would
>not refuse to sell to Sun customers were Sun to have AES in
>its portfolio. That being said, they might have been less
>enthusiastic in their support of Sun if Sun had started to
>compete with them in the late 1990ies.

By the early 2000s Oracle was less enthusiastic in their support
of Sun.

>Which they do. Only they do not sell a database. It's easy
>to say with 20/20 hindsight they should have had a database
>since at least 2000, but remember that by that time Oracle
>had largely won the battle, so buying any of the remaining
>losers might not have revived its chances.

Have Oracle and IBM won the RDBMS marketshare battle or war?
Your use of battle suggests that you agree that the situation
is fluid.

>feel. I know of a large organisation that uses lots of large
>and medium SPARC servers running Solaris, but will not consider
>Solaris/x86 for the ever increasing horde of low-end x86 servers,
>where they mandate Linux. Go figure.

Why not?

>That's something no-one can answer. It's better than Oracle in
>many ways, but since MS-DOS and Windows we know that technical
>excellence means diddly squat.

Does Oracle and IBM have a monopoly in marketoons and salescritters
capable of selling RDBMS?

>I don't follow IBM's DB2 pricing on Solaris. Do they lower the
>price of DB/2 if people by a P-series?

For academia, I believe so.

>Still, this does not mean that Sun offering AES for a lower fee
>would necessarily stop people from buying Oracle, or DB/2.

What would Sun need to do gain your DBMS business?

John
groenveld@acm.org

In article <20050610141716.0f0071fe.tengo@DELETEMEecc.lu>,
Stefaan A Eeckels <tengo@DELETEMEecc.lu> wrote:
>Indeed. Unfortunately, AES is not widely perceived as
>"a big RDBMS", and Sun cannot afford Oracle.

I'm no MBA, but a quick read of Al Ries and Jack Trout's book
suggests that problems of perceptions can be surmounted.
<URL:http://www.amazon.com/exec/obidos/ISBN=0887306667>

John
groenveld@acm.org

In article <slrndaik8l.cm8.comp.unix.solaris@usenet.andreas-borchert.de>,
Andreas F. Borchert <comp.unix.solaris@expires-on-2005-06-18.usenet.andreas-borchert.de
> wrote:
>On 2005-06-08, John D Groenveld <groenvel@cse.psu.edu> wrote:
>> Can't wait to see Sun's mobile workstation system.
>
>Once upon a time, there was a SPARCstation Voyager [1].

And if I am right, this was the first machine ever with a
suspend to disk feature.

--
EMail:joerg@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
js@cs.tu-berlin.de (uni)
schilling@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
URL: http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

John D Groenveld <groenvel@cse.psu.edu> wrote:
> In article <20050609223241.769490c2.tengo@DELETEMEecc.lu>,
> Stefaan A Eeckels <tengo@DELETEMEecc.lu> wrote:
>>Are you suggesting that Sun should have been more like IBM
>>proposing its own integrated service and software? I can
>
> Yes.
>
>>understand that going head-to-head with Oracle (which sold
>>a lot of systems for Sun) during the boom would have been
>>considered foolish (especially by Wall Street pundits).
>
> Why isn't IBM considered foolish by pundits for competing with
> Oracle by selling DB2 on IBM, Sun, and other company's systems?

I just have to jump in here.

I remember reading an investment book a few years ago, which talked about
investing in IBM. IBM had consistently underperformed for decades, and yet
they were still a blue chip stock (A 'big blue' chip stock in fact! Hah!).
The reasoning went that they were KNOWN to be a good company, and no broker
would ever get in trouble for recommending their stock. If they
underperformed, it was 'just one of those things,' regardless of the fact
that they were consistently underperforming.

To some degree, IBM selling DB2 is the same thing--they're such a big and
OLD company in the computing industry, that anything they do which would
be scorned from another company is accepted as being part of their overall
(implied good) strategy.

Stefaan A Eeckels <tengo@DELETEMEecc.lu> wrote:
> On Fri, 10 Jun 2005 08:15:15 +0000 (UTC)
> Przemek Bak <przemolicc@poczta.fm.-no-spam.invalid> wrote:
>
> > But Sun could have better portfolio and could better integrate
> > everything.
> > Do you want to have big RDBMS ?
> > Go to Sun.
> > Do you want to have big server ?
> > Go to Sun.
> > Do you want to have big storage ?
> > Go to Sun.
> > Do you want to have big RDBMS on big server with big storage ?
> > Go to Sun.
>
> Indeed. Unfortunately, AES is not widely perceived as
> "a big RDBMS" [...]

It's a matter of marketing. Don't you think ? ;-)

przemol

On Mon, 13 Jun 2005 05:45:11 +0000 (UTC)
Przemek Bak <przemolicc@poczta.fm.-no-spam.invalid> wrote:

> Stefaan A Eeckels <tengo@DELETEMEecc.lu> wrote:
> > On Fri, 10 Jun 2005 08:15:15 +0000 (UTC)
> > Przemek Bak <przemolicc@poczta.fm.-no-spam.invalid> wrote:
> >
> > > But Sun could have better portfolio and could better integrate
> > > everything.
> > > Do you want to have big RDBMS ?
> > > Go to Sun.
> > > Do you want to have big server ?
> > > Go to Sun.
> > > Do you want to have big storage ?
> > > Go to Sun.
> > > Do you want to have big RDBMS on big server with big storage ?
> > > Go to Sun.
> >
> > Indeed. Unfortunately, AES is not widely perceived as
> > "a big RDBMS" [...]
>
> It's a matter of marketing. Don't you think ? ;-)

It _was_ a matter of marketing. Picking up AES today, and trying
to convince Oracle users to switch looks like a solid imitation
of Don Quixote fighting the windmills. Especially given the
legendary quality of Sun's marketing droids.

Should Sun have written/acquired a DBMS when it transitioned
from a workstation company to a server company? In those days,
Sun's DBMS vendor neutrality was seen as a good thing, and
hindsight is obviously always 100% accurate.

--
Stefaan
--
As complexity rises, precise statements lose meaning,
and meaningful statements lose precision. -- Lotfi Zadeh