This message is cross-posted to 2 groups because of making comparsons.

When I review SQL Server 2005 functions, and get to know languages
called MDX, DMX, XML/A. MDX is kinda dimenstional version of SQL, which
Oracle seems to try to include as extension to SQL such as "group by
rollup", "group by cube", etc., and looks not as powerful as DMX. DMX
is kinda data mining version of SQL, which I don't know Oracle's
counterpart, but I heard about the industry standard called DMML, if I
am correct. XML/A is the protocol between SSAS and it's clients, I am
not sure whether it's a language people use, but I know the structural
nature of XML is quite useful for, say, storing metadata kinda things.
Anyone can make comparison between Oracle and Microsoft on this aspect?

On the other hand, Oracle embraces Java such as in Java stored
procedure, SQLJ, JDBC, but Microsoft's kinda hate it (such as in that
you have to download JRE by yourself, no garantee the version will be
correct). Any comments too?

Thanks in advance.

<dino_hsu_1019 (AT) yahoo (DOT) com> wrote

question is what business problem are you trying to solve?
Jim

Aside from the SQL capability that Oracle has provided for compound
rollup, cube, grouping sets and windowing functions (which are
included in baseline Oracle) you can also wander around the Data Mining
docco and the OLAP docco at
http://www.oracle.com/pls/db102/port..._db?selected=3


Data Mining Administrator's Guide
Data Mining Application Developer's Guide
Data Mining Concepts
Data Mining Java API Reference (Javadoc)

OLAP Analytic Workspace Java API Reference (Javadoc)
OLAP Application Developer's Guide
OLAP Developer's Guide to the OLAP API
OLAP DML Reference
OLAP Java API Reference (Javadoc)
OLAP Reference


--
Hans Forbrich
Canada-wide Oracle training and consulting
mailto: Fuzzy.GreyBeard_at_gmail.com
*** Top posting [replies] guarantees I won't respond. ***

My business problems are reporting, OLAP, data mining, where OLAP is
the most critical because we already have operational reporting in
place, it's not going to change too soon, data mining is yet to prove
its use, OLAP is the area where we can make a big progress.
Oracle is our Enterprise standard, so the transaction db can't be
changed. I am thinking about adopting MSFT's BI capabilities for its
relative low price and good functions.
I am asking the camparison of BI languages between Oracle and MSFT,
it's very sepcific to any business problem, but the understanding is
useful in evaluating and learning them.
As for the Java part, I have just figured out a solution of combining
MQ series plus Java language plus Oracle stored procedure, saving the
cost of Oracle Procedure Gateway to MQ. In the process, I have noticed
the different Java strategy applied by Oracle and Microsoft.

My business problems are reporting, OLAP, data mining, where OLAP is
the most critical because we already have operational reporting in
place, it's not going to change too soon, data mining is yet to prove
its use, OLAP is the area where we can make a big progress.

Oracle is our Enterprise standard, so the transaction db can't be
changed. I am thinking about adopting MSFT's BI capabilities for its
relative low price and good functions.
I am asking the camparison of BI languages between Oracle and MSFT,
it's not very sepcific to any business problem, but the understanding
is
useful in evaluating and learning them.

As for the Java part, I have just figured out a solution of combining
MQ series plus Java language plus Oracle stored procedure, saving the
cost of Oracle Procedure Gateway to MQ. In the process, I have noticed
the different Java strategy applied by Oracle and Microsoft.

On 19 Dec 2005 20:40:03 -0800, "Dino Hsu" <dino_hsu_1019 (AT) yahoo (DOT) com>
wrote:

Oracle has an OLAP product called Discoverer. Why include another
level of complexity in the picture?

--
Sybrand Bakker, Senior Oracle DBA

I have looked at Oracle's solution not long ago, including OLAP option
in db, OWB in developer, Discoverer application server, etc.. I can't
remember all of them. I will not try to answer why not, I will try why
yes.

We have been using Cognos' Power Play 6.5 and Impromptu 5.0, and IBM
DB2 Intelligent Miner, all of them quite expensive, and as BI is
overlooked in the past at our enterprise level, we can't, as a
medium-sized local market, offord to pay the MA for them. I am
attracted by good features of SQL Server 2005, for example, custom
rollup and semi-additive measures. You know what? It really makes BIG
difference with / without OLAP flexibitiy. Reporting service is also
interesting, plus data mining, it's of very low cost indeed. A SQL
Server server + 25 CAL license is about one fourth (2XCPU) to one
eighth (4XCPU) that of IBM DB2 IM per-processor license, and with a lot
more overall functions.

Of course, I don't mean SQL Server 2005's BI is perfect, (I am still
verifying by studying) the issue is it has limited front-end tools, but
the trend looks promising.

Dino Hsu wrote:

Well if you can use the trend, others call if vapourware, to do your
reporting right this minute by all means do so. I usually find that
the currently installed toolset works better.

But hey ... what could possibly be wrong with a version 1.0 product
from Microsoft. ;-)
--
Daniel A. Morgan
http://www.psoug.org
damorgan@x.washington.edu
(replace x with u to respond)

The approach with SQL Server Data Mining is to provide familiar modeling
concepts and APIs so that it's easy for database developers develop
solutions and embed data mining.

The DMX language uses standard database paradigms for creating (CREATE
MINING MODEL ..), training (INSERT INTO <model>) and querying (SELECT ...
FROM <model> PREDICTION JOIN <data>) mining models as standard database
objects. In addition, developers can connect to and query models using
standard object models like OLE DB, ADO and their .NET/managed counterparts,
ADO.NET and ADOMD.NET.

PMML (the XML format for interoperability with other data mining products)
is also supported so third-party mining models can be loaded and queried by
SQL Server 2005 as well.

Another facet of the API story is the ability for third-party developers to
plug in their own data mining algorithms and viewers into SQL Server Data
Mining, at the same level as the built-in algorithms and viewers.

You can find a lot of content from the development team, including technical
whitepapers, articles and live samples, at our web site:
http://www.sqlserverdatamining.com (no marketing info here).

--
-Raman Iyer
SQL Server Data Mining
[This posting is provided "AS IS" with no warranties, and confers no
rights.]

<dino_hsu_1019 (AT) yahoo (DOT) com> wrote

yes Analysis Services blows Oracle out of the water.

i mean.. www.olapreport.com do you see oracles' market share?
do you see microsofts?

Excel against OLAP is the most wonderful and exciting technology in the
world
congrats!