Thought you could not find anything on the web?!?
You're just suffering from selective reading, correct?
As any zealot would.

Well, I don't trust that bit, as it's a MS supported/bought group.
Anything I've read from them favors Microsoft, as some reports are
actually sponsored, or paid by MS.

I have used log shipping and it is prone to problems. You have to basically
freeze the database (source) and send it to the target.  You have to be
careful applying logs, SS doesn't seem to be able to know what order to
apply logs in. (which seems rather silly, it should know the order)
Jim

Thus proving that the difference, if any, is insignificant. Thanks.

Yes

No - not with the Standard Editions

Note that the pricing structures quoted in the Register is for
multicores machines that SQLServer doesn't run on - such as the IBM
hardware, and also the Niagra. There are no 11 core machines

1 minute of Google gave me this :
http://www.oracle.com/database/product_editions.html

Clustering is included in Standard Edition, but not in Standard Edition One.

HTH
Matthias Hoys

Recent events in the US has meant that quite a few sites have failed
over to their standby environments for much longer than 30 days. If you
have had any sort of US based bank account in the last 5 years or so,
you have been a customer.

I sent you the list already.

Oracle doesn't list what is in Standard (as that list would be huge),
they list the differences. Clustering a la MS style is available for all
editions of Oracle. Clustering a la Oracle (RAC) is not available for
Standard Edition One, provided for free in Standard Edition, and an
optional extra with Enterprise Edition

Finally something we can agree on

Yes it does say it there. What part are you confused by ?

Tony - this is known to most of the people on this group already, so
this discussion will be anathema to them, but for completeness, let me
explain what is going on

Both SQLServer and Oracle have multiple data engines in their products -
a SQL engine, and an OLAP engine.

Oracle's SQL engine is stronger than SQLServers SQL engine in terms of
BI, in that there are (standard, by the way) SQL Extensions that support
BI activities such as analysis, aggregation and modelling.

See
http://download-west.oracle.com/docs/cd/B19306_01/server.102/b14223/aggreg.htm#i
1007462
http://download-west.oracle.com/docs/cd/B19306_01/server.102/b14223/analysis.htm
#i1007779
and
http://download-west.oracle.com/docs/cd/B19306_01/server.102/b14223/sqlmodel.htm
#sthref1855
for a discussion of what these are.

In case you don't have an OTN account, I've summarized these
capabilities below

ROLLUP Extension to GROUP BY
CUBE Extension to GROUP BY
GROUPING Functions
GROUPING SETS Expression
Ranking Functions
Windowing Aggregate Functions
Reporting Aggregate Functions
LAG/LEAD Functions
FIRST/LAST Functions
Inverse Percentile Functions
Hypothetical Rank and Distribution Functions
Linear Regression Functions
Linear Algebra
Frequent Itemsets
WIDTH_BUCKET Function
User-Defined Aggregate Functions
Data Densification for Reporting
Time Series Calculations on Densified Data

These BI capabilities in the SQL engine are in ALL editions of Oracle

SQLServer doesn't have many of these equivalent capabilities in the SQL
engine, so SQLServer users tend to go to the OLAP engine in SQLServer
for equivalents.

The same is not true for Oracle users. The Oracle OLAP engine provides a
whole new level of BI capabilities over and above what the Oracle SQL
engine provides. Some of these capabilities overlap with what the
SQLServer OLAP engine provides. So it sort of looks like this

Microsoft <---SQL---><--OLAP-->
Oracle    <-------SQL------><----OLAP---->

Oracle charges for it's OLAP capabilities, SQLServer does not. But to
characterize Oracle as not having BI capabilities in all editions of the
database is misleading.

What about you take this bloody trolling off-line.

When are you going to get it through your thick skull I am not paid by
anybody to do any bidding anywhere.

The only income I get is a) from consultancy through clients b) through
development for (a) and c) events I run through my user group.

I get no income from Microsoft - get it into your thick arrogant skull,
don't be so dam ignorant.

So you expect to post crap and lies about other vendor products without
challenge?

The only troll here pal is you and how you manage to extract $35 out of
people to be members of a free community I'd love to know (and understand)!

And what comments would those be? Making stuff up ->again<- are we Daniel?

I'm putting him in my kill file.  (to paraphrase a certain English man) He
is just noise signifying nothing.
Jim

What level of ANSI 92 or didn't you know there are different levels?  That
is the version not the level.  SS isn't completely ANSI 92 compliant at all
levels. (no one is)
Jim

Tony perhaps it hasn't occured to you that the level of expertise and
competence separating Oracle DBAs with SQL Server DBAs is vast.

Number of SQL Server DBAs that need to know more than 1 operating
system? Zero!

Number of SQL Server DBAs managing line-of-business applications in a
Fortune 500 environment? Near Zero!

Number of SQL Server DBAs that can write a perl, bash, or korn shell
script? Near Zero!

Number of SQL Server DBAs that have more than 15 years experience in
relational database architecture, concepts, and design? Near Zero!

If you don't have the skill set you shouldn't expect the pay. If
organizations want the tool set Oracle provides they should expect
to pay for the value received. I make no apology for either my bill
rate or my competencies.

A company that wishes to save money by purchasing SQL Server might
want to look over its shoulder at SarbOx and Basel II audit
requirements. Penny-wise ... Pound foolish.

I did and I couldn't find the citation, which is why I asked you for it.
I want to check what level of ANSI compliance MS claims, to validate
your claim that their implementation of SQL is somehow more the ANSI
standard compliant than Oracle's. Note that Oracle does this at
http://download-west.oracle.com/docs/cd/B19306_01/server.102/b14200/ap_standard_
sql003.htm#i7719

I'm looking for the MS corresponding compliance statement. Can you
provide a URL ?

I did. The page references I gave you provide the following

        Min    MAX
Oracle        39,363    46,026
SQL Server    36,128    40,846

I'm not sure where you are getting some of your figures from

The math is as follows

Difference    8.95%    12.68%

In the previous post, I did make a mistake in the maths with the low
end. So the range is 9-13% higher than SQL Server ranges. This is a
higher, bit is not substantially higher. And as I indicated, I believe
that this is because, on average, the Oracle DBAs are managing systems
with more mission critical requirements.

Please do. Feel free to bring your own bias to the party. We have
admired your constraint in not doing so so far.

Thank you for encouraging the young 'uns in the correct direction.  The
last time I was officially classified as a DBA, about 5 years ago, I
made 25% more than your stated Max, plus overtime.  I believe it was
about a third of the highest paid DBA in that large shop.

I made a conscious decision to bias towards Oracle/unix in 1989, and it
was a good one.  Haven't yet seen a reason to change.  Sometimes, you
do get what you pay for, especially for industrial strength solutions.
A lot of my work has been throwing out stuff, particularly MS, that
doesn't scale well, fixing stuff that was not done right, yes, even in
Oracle, and making cross-db stuff work much better in Oracle.

But I must disagree with those who say apps will be thrown out in 3-5
years, probably because I'm skewed towards mission-critical
manufacturing and financials.  Seems people figure out early on that
Oracle is better, or they won't even be talking to me.  Better even for
LCD cross-db apps.

jg
--
@home.com is bogus.
Interesting article about transaction isolation levels:
http://www.oracle.com/technology/oramag/oracle/05-nov/o65asktom.html

During the days I was into mainframe programming, research (not by me,
but independent) showed the average life time of an application was
5 years.

I cannot see how, or why this cycle would have decreased over the years.

So, I can only agree with what Hans says, even if the experience
is old. That could explain my 5, versus Hans' 3 years span.

On the other hand, I do agree there's a substantial portion
"not invented here" syndrome going on, causing new
development not to choose the best tool, but best cost/benefit
ratio on the short term (i.e. one or two years, not the lifespan
of the application built).
And of course, (human) nature resists against change. What ever change,
for better or for worse.

What I have seen is shops switching applications in 3 years - or (as Frank
points out) 5 to 7 years.  My observation around this is:

- switch because of changes in functionality
- switch vendors, often to one with different infrastructure experience
 (we've built this amazing thing in DB2 or SQL Server or Oracle)
- the shop often wants it rewritten to meet the shop's "standard"
- we use DB2, you have to make it available in DB2 even though it was
 written for Oracle
- the reason is the administrators (database, system,network)
- therefore, to appease the administrators we will accept an inferior port
 of to retain the apparent improvement in functionality,
- unfortunately, the inferior port tends to have performance issues
 which leads to early review of replacement strategy.

And yes, I have seen shops switch vendors 'on a dime' and commit to the
vendor extensions - when there is a management endorsed exit strategy.  

It means the team is committed to working on that project, from initial
development through retirement, for the duration.  It means using true
cost techniques that include Capital Expenses as well as Operation
Expenses over the lifetime.  And it means measuring the return on
investment and using [part of] the savings that result from the return as
capital for the next project rather than sluffing it to the general profit
pool.

A pleasant thought but not borne out by statistics gathered during
numerous research studies. Put SAP on top of DB2 and you are staying
there for several generations. Same goes with any other combination.

If is always enjoyable to see a geek try to turn into a marketer.