Comments in-line.

Niall Litchfield wrote:

>Most of your comments are along the lines of ... but that feature isn't any
>help to me ... which is fair enough but Oracle having useless features for
>years isn't really a winning argument to me. Specific comments embedded.
>
Well .... uh ... ummm ... uhh ... yes.

>><snipped>
>>
>>If so then I submit that SQL, PL/SQL, SQLJ, Java, C, C++, Cobol,
>>Fortran, Perl, and
>>a plethora of scripting languages (PHP, KSH, etc.) is as close as I need
>>to be.
>>
>>
>
>I tend to agree - this probably does't surprise you - but if there is a case
>for having stored procedures written in java in the db, then probably the
>ability to write stored procedures in C#,VB,C++,J# (probably the most likely
>order that they would be used) would appear to be an improvement on the
>Oracle offering.
>
You are aware, I hope, that procedures and functions in Oracle can now
be compiled into C. The
fact that it isn't the Microsoft branded products is hardly an issue as
most Oracle is not run on a
Microsoft operating system.

>>>enhanced select TOP ....
>>>
>>>
>>I'd need a more complete description to respond.
>>
>>
>
>does Oracle offer select TOP .. at all? One of those easily worked around
>limitations that probably irritate a fair number of people trying to write
>db independent code - lets not go down the is this a good idea or not route.
>
Whenever I see TOP I thnk of the UNIX utility. Are you talking about
MAX? GREATEST? CEIL? Or
are you referring to the ability to grab the "top 5" that match some
criterion.

If the later ... there is no built-in function. But 10 seconds with an
in-line view and it can easily be done.

>>>seperate date and time datatypes?
>>>
>>>
>>>
>>Why? To what possible gain? Time bereft of date is almost always
>>meaningless.
>>
>>
>
>We have a report that we need to write on activity in working hours and out
>of working hours, we query on just the time part of the field. Again
>relatively easy but a nice to have feature
>
So use 01-JAN-1900. Ok I do get your point. But I'd argue that activites
can span days and that it
is the rare exception when that doesn't happen.

>>>try..catch in PL/SQL?
>>>
>>>
>>>
>>Explain.
>>
>>
>
>Create or replace longProcedure as
>..
>begin
>try
>query that might return no rows
>catch NO_DATA_FOUND
>....
>end try
>
>do some stuff
>try
>query that might throw a user defined exception but not normally
>NO_DATA_FOUND
>catch UDE
>...
>end try
>
>...
>EXCEPTION
>WHEN NO_DATA_FOUND
>... I have really had an error here.
>
>end ;
>/
>
>in other words if you expect certain errors in certain parts of code then
>deal with them there. Now personally i'd probably write that as
>Create or replace longProcedure as
>begin
>call subProc1
>call subProc2
>EXCEPTION
>end;
>
>etc
>
I would too as it would modularize my code, make development and testing
easier, allow for
code reuse, etc. But I do your CATCH stuff all the time like this:

BEGIN
BEGIN
... stuff
EXCEPTION
WHEN some_user_defined_exception THEN
... handle it
END;

... repeat ad-infinitum

EXCEPTION
WHEN OTHERS THEN
... bail out gracefully
END;
/

>>Now I'm completely confused by what you are asking. Unfortunately that
>>didn't stop me from responding as best I could. ;-)
>>
>>
>
>My only point was that the OP stated that the article listed as new a whole
>bunch of features that Oracle had had 'for years' and yet when you got into
>specific features rather than marketing waffle a fair number of them seemed
>new both to mssql and to oracle. I have no problem arguing that oracle is a
>technically superior product - though I have issues with its pricing and
>marketing - but if we are going to do conparisons lets do them based on
>facts and not broad generalisations. One might as well look at the marketing
>for 10g and say 'ah but sqlserver has been easy to manage for years'.
>
>
>
You'd be correct. SQL Server has been easy to manage for years.
Partially one must admit
if one is honest because those that use it rarely have to meet the level
of demanding apps
that one routinely puts on top of Oracle. Just for one example, no one
would in their wildest
imagination have tried to host Amazon.com on SQL Server. I just finished
a project with a
small telephone company that pushed SQL Server to the point that it
croaked. We went to
Oracle and even without tuning handled 10X the throughput.

So if you have a half-way decent product but you never really push it
too hard you can
always get by without tuning because no one expects very much. Most of
what I've seen
done in SQL Server, put into Oracle or DB2 or Informix could also not
require much DBA
support because even a badly tuned database/instance would be adequate.

Is Oracle one of the most incompetent marketing organizations on the
planet for its size?
Absolutely. When a room full of industry professionals (my students) can
name a dozen or
more Microsoft products and still think Oracle is a one product company
Oracle's marketing
staff should be reduce to half-pay until they earn the rest of their
salary. Pricing wise ... its
going to change. Because if it doesn't Oracle's marketshare will erode.
That is the law. The
only way Oracle can continue its current pricing is not to offer more
candy on the CD. But
to make people aware of the fact that the candy exists and how to use
it. Something they
have failed at time and time again.

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

"Daniel Morgan" <damorgan@x.washington.edu> wrote in message
news:1067277048.821099@yasure...
> You'd be correct. SQL Server has been easy to manage for years.
> Partially one must admit
> if one is honest because those that use it rarely have to meet the level
> of demanding apps
> that one routinely puts on top of Oracle. Just for one example, no one
> would in their wildest
> imagination have tried to host Amazon.com on SQL Server. I just finished
> a project with a
> small telephone company that pushed SQL Server to the point that it
> croaked. We went to
> Oracle and even without tuning handled 10X the throughput.



A UK example but Tescos accounts for 1 in 8GBP of all UK spending and
tesco.com beats the hell out of their competitors.

http://www.microsoft.com/resources/c...aseStudyID=130
66

It may be that amazon.com requires the power of Oracle but if ( say)
walmart.com doesn't and tescos is larger.

--
Niall Litchfield
Oracle DBA
Audit Commission UK
*****************************************
Please include version and platform
and SQL where applicable
It makes life easier and increases the
likelihood of a good answer
******************************************











>
> So if you have a half-way decent product but you never really push it
> too hard you can
> always get by without tuning because no one expects very much. Most of
> what I've seen
> done in SQL Server, put into Oracle or DB2 or Informix could also not
> require much DBA
> support because even a badly tuned database/instance would be adequate.
>
> Is Oracle one of the most incompetent marketing organizations on the
> planet for its size?
> Absolutely. When a room full of industry professionals (my students) can
> name a dozen or
> more Microsoft products and still think Oracle is a one product company
> Oracle's marketing
> staff should be reduce to half-pay until they earn the rest of their
> salary. Pricing wise ... its
> going to change. Because if it doesn't Oracle's marketshare will erode.
> That is the law. The
> only way Oracle can continue its current pricing is not to offer more
> candy on the CD. But
> to make people aware of the fact that the candy exists and how to use
> it. Something they
> have failed at time and time again.
>
> --
> Daniel Morgan
> http://www.outreach.washington.edu/e...ad/oad_crs.asp
> http://www.outreach.washington.edu/e...oa/aoa_crs.asp
> damorgan@x.washington.edu
> (replace 'x' with a 'u' to reply)
>
>

Daniel Morgan <damorgan@x.washington.edu> wrote in message news:<1067277048.821099@yasure>...

> >>>seperate date and time datatypes?
> >>>
> >>Why? To what possible gain? Time bereft of date is almost always
> >>meaningless.

Don't know whose comment this was, but what about DATE bereft of TIME?
Also DATE, TIME and TIMESTAMP data types are part of the SQL standard.

> When a room full of industry professionals (my students) can
> name a dozen or more Microsoft products and still think Oracle is a one
> product company Oracle's marketing staff should be reduce to half-pay until
> they earn the rest of their salary. Pricing wise ... its
> going to change. Because if it doesn't Oracle's marketshare will erode.

Does 70% of Oracle's turnover come from DBMS sales or is it now only a
little over 60%? I forget. The point is that Oracle *is* unhealthily
dependent on database sales. And yes they are overpriced, but because
of the heavy dependence on database sales, if they reduce the prices
they hit the bottom line very, very hard. But if they don't they keep
losing share, as you say. Meanwhile IBM and Microsoft sell their
products at a fraction of the cost. I'm not a financial analyst but to
me it seems a nasty position.


DG

Database Guy wrote:

>Daniel Morgan <damorgan@x.washington.edu> wrote in message news:<1067277048.821099@yasure>...
>
>
>
>>>>>seperate date and time datatypes?
>>>>>
>>>>>
>>>>>
>>>>Why? To what possible gain? Time bereft of date is almost always
>>>>meaningless.
>>>>
>>>>
>
>Don't know whose comment this was, but what about DATE bereft of TIME?
>Also DATE, TIME and TIMESTAMP data types are part of the SQL standard.
>
It was mine.

>>When a room full of industry professionals (my students) can
>>name a dozen or more Microsoft products and still think Oracle is a one
>>product company Oracle's marketing staff should be reduce to half-pay until
>>they earn the rest of their salary. Pricing wise ... its
>>going to change. Because if it doesn't Oracle's marketshare will erode.
>>
>>
>
>Does 70% of Oracle's turnover come from DBMS sales or is it now only a
>little over 60%? I forget. The point is that Oracle *is* unhealthily
>dependent on database sales. And yes they are overpriced, but because
>of the heavy dependence on database sales, if they reduce the prices
>they hit the bottom line very, very hard. But if they don't they keep
>losing share, as you say. Meanwhile IBM and Microsoft sell their
>products at a fraction of the cost. I'm not a financial analyst but to
>me it seems a nasty position.
>
>
>DG
>
>
Exactly. Have you ever seen the huge list of Oracle products? I'll bet
no one here can name
half of them. And I include Jonathan and Richard and Howard in that "no
one." It is an
insane situation. So the corporation is primarily dependent for its
revenues on a single product
that is almost always marketed like it was MS Access on steroids. So
even the full feature set
of the database is almost unknown.

There is no question that databases are becoming a commodity. Prices
must and will fall. And
unless Oracle replaces its current marketing zeros with a suite full of
ones I expect things will
get a bit tighter for everyone in Redwood Shores.

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

Daniel Morgan <damorgan@x.washington.edu> wrote in message news:<1067277048.821099@yasure>...
>
> Is Oracle one of the most incompetent marketing organizations on the
> planet for its size?
> Absolutely. When a room full of industry professionals (my students) can
> name a dozen or
> more Microsoft products and still think Oracle is a one product company
> Oracle's marketing
> staff should be reduce to half-pay until they earn the rest of their
> salary.

How true. I've maintained this for years. Back in the days
when Ingres was alive and claimed at every drop of a hat that
Oracle could sell only because it had "good marketing", I simply
had to laugh! Oracle NEVER had good marketing. Ever!

That is if we consider "marketing" in the conventional sense
as the science of creating long term demand for a range of products.

Of course, others call "marketing" to what used to be called PR
and advertising. In those areas, sure, Oracle might do
better than many.

But not in real marketing: they don't have the foggiest.

> candy on the CD. But
> to make people aware of the fact that the candy exists and how to use
> it. Something they
> have failed at time and time again.

Exactly. And making it cheaper for the same candy ain't gonna
solve the problem.

Cheers
Nuno Souto
wizofoz2k@yahoo.com.au.nospam

--
"Database Guy" <dbguy101@hotmail.com> wrote in message
news:7fdee71c.0310271703.11557ae0@posting.google.c om...
> Daniel Morgan <damorgan@x.washington.edu> wrote in message
news:<1067277048.821099@yasure>...
>
> > >>>seperate date and time datatypes?
> > >>>
> > >>Why? To what possible gain? Time bereft of date is almost always
> > >>meaningless.
>
> Don't know whose comment this was, but what about DATE bereft of TIME?
> Also DATE, TIME and TIMESTAMP data types are part of the SQL standard.
>
> > When a room full of industry professionals (my students) can
> > name a dozen or more Microsoft products and still think Oracle is a one
> > product company Oracle's marketing staff should be reduce to half-pay
until
> > they earn the rest of their salary. Pricing wise ... its
> > going to change. Because if it doesn't Oracle's marketshare will erode.
>
> Does 70% of Oracle's turnover come from DBMS sales or is it now only a
> little over 60%? I forget. The point is that Oracle *is* unhealthily
> dependent on database sales. And yes they are overpriced, but because
> of the heavy dependence on database sales, if they reduce the prices
> they hit the bottom line very, very hard. But if they don't they keep
> losing share, as you say. Meanwhile IBM and Microsoft sell their
> products at a fraction of the cost. I'm not a financial analyst but to
> me it seems a nasty position.
>

The price difference was true at one point (except you didn't get nearly the
same features as today in Sql Server - so it was worth a lot less), but I
don't see where the price is that much different any more, if you compare
(as one example) the Standard Edition of Oracle to the equivalent
(non-Enterprise) version of Sql Server.

For example we bought a copy of Sql Server 3 years ago for 5 users at $1,500
.. That is the same price for the Standard Edition of Oracle today.

And, now Oracle has a new price for basically the standard edition of
$195/user. For more info, see URL
http://www.infoworld.com/article/03/...lesmbdb_1.html

>
> DG

On Mon, 27 Oct 2003 21:51:31 +0000, Niall Litchfield wrote:

Morning Niall,

>
> A UK example but Tescos accounts for 1 in 8GBP of all UK spending and
> tesco.com beats the hell out of their competitors.
>
> http://www.microsoft.com/resources/c...aseStudyID=130
> 66
>

Ummm - Tesco's use Oracle because a friend of my wife works there !


> It may be that amazon.com requires the power of Oracle but if ( say)
> walmart.com doesn't and tescos is larger.
>

Amazon run their business on HP-UX servers (or so I've been told by HP
anyway) and SQL Server doesn't run on HP-UX. I suspect this limited the
choice of database somewhat.

MS need to get their head out of the sand and port their products to
other/competing operating systems if they want to 'expand'. Even
Interbase/Firebird runs on Windows/Unix/Linux etc and it's hardly a SQL
Server. (It actually better - IMHO)

Cheers,
Norm.


--
Delete the obvious bit from my email address to reply by email.

"Database Guy" <dbguy101@hotmail.com> wrote in message news:7fdee71c.0310271703.11557ae0@posting.google.c om...

> little over 60%? I forget. The point is that Oracle *is* unhealthily
> dependent on database sales. And yes they are overpriced, but because
> of the heavy dependence on database sales, if they reduce the prices
> they hit the bottom line very, very hard. But if they don't they keep
> losing share, as you say. Meanwhile IBM and Microsoft sell their
> products at a fraction of the cost. I'm not a financial analyst but to
> me it seems a nasty position.
>

and it's gonna get worse. just wait for M$ to bundle the db with the OS...
--
Cheers
Nuno Souto
wizofoz2k@yahoo.com.au.nospam

"Norman Dunbar" <Norman@RE-MO-VE.BountifulSolutions.co.uk> wrote in message
newsan.2003.10.28.09.18.05.358304@RE-MO-VE.BountifulSolutions.co.uk...
> On Mon, 27 Oct 2003 21:51:31 +0000, Niall Litchfield wrote:
>
> Morning Niall,
>
> >
> > A UK example but Tescos accounts for 1 in 8GBP of all UK spending and
> > tesco.com beats the hell out of their competitors.
> >
> >
http://www.microsoft.com/resources/c...aseStudyID=130
> > 66
> >
>
> Ummm - Tesco's use Oracle because a friend of my wife works there !

I'm sure that Tescos are a Oracle customer, but it seems unlikely to me that
ms would publish this case study if it were untrue.

Cheers

Niall

"Burt Peltier" <burttemp1ReMoVeThIs@bellsouth.net> wrote in message
news:x1onb.55753$h47.51947@bignews4.bellsouth.net. ..
> The price difference was true at one point (except you didn't get nearly
the
> same features as today in Sql Server - so it was worth a lot less), but I
> don't see where the price is that much different any more, if you compare
> (as one example) the Standard Edition of Oracle to the equivalent
> (non-Enterprise) version of Sql Server.
>
> For example we bought a copy of Sql Server 3 years ago for 5 users at
$1,500
> . That is the same price for the Standard Edition of Oracle today.
>
> And, now Oracle has a new price for basically the standard edition of
> $195/user. For more info, see URL
> http://www.infoworld.com/article/03/...lesmbdb_1.html

Std Edition one is IMO exactly the right price for Std Edition. I really,
really hate the idea of running any database - especially one aimed at
running a small business or large corporate workgroup on a single processor
machine. Still it has been at least 2 years since Larry announced new
pricing so we are about due a new pricing policy for all the products.


--
Niall Litchfield
Oracle DBA
Audit Commission Uk