Information designer is a new tool in BO 4.0 that features
the creation of universes (.unv, .unx) with multiple data sources, which is not
available in previous versions of Business Objects.
IDT follows layered approach. The layers are
• Connection
Layer.
• Data
Foundation Layer
• Business
Layer
In the connection layer we create the connection string
(.cnx) to the data sources. Each data source will have one connection string.
In the Data foundation layer we will create the data
foundation file (.dfx) with one or multiple connections. That means we will be able to bring the tables
from multiple the data sources. This layer will contain the joins, contexts,
derived tables, alias etc.
Eg: we will be able to access the tables from both SQL and
Oracle in a single data foundation.
In the Business layer we will be able to create the file
(.blx) will contain the user layer,
including objects used to access the data. We publish this business layer and
use it as a Universe. So we will be able to create multiple universes on a
single data foundation, which will save lot of time.
Here is the pictorial representation of IDT.
Here I am listing some of the pros and cons of IDT.
Pros:
- Building one data
foundation with all data sources and build universes on top of it eliminates
the creation of linked universes
- Can build views on top of
Business layer, so that we can have security privileges easily.
- We can create Dimensional
(OLAP) universes that support OLAP dimensions and hierarchies
- In earlier version (3.1)
if we have a context then every join must be a member of one context. In
4.0, only the joins that are part of a loop need to be in contexts. All
other joins will be equally accessible to all contexts.
- New search feature enables
to search tables based on
- Table Name
- Column Name
- Family
- Context
- We can add tables too
family and color code them. You can import and export those families.
- Ability to add calculated
columns to the Tables in data foundation so it can be used in multiple
business layers
- When creating the
calculated column using the tables from to databases (SQL, Oracle) IDT
provides Business objects SQL, so that it is common to both DB’s and will
be translated to SQL during runtime.
- Developing universes in IDT significantly lower TCO
(Total Cost of Ownership) for your universes
- In IDT we can fully test
as we go, which helps to change joins, conditions etc.
Cons:
- No backward compatibility i.e Universes built in
IDT cannot be opened or edited in Universe Design Tool.
- We have
to format the objects individually which is cumbersome.
- If we
define a custom format we cannot use the format for other objects. We have to
define the format for every object
- No
ability to change dimension to an attribute (detail) or vice versa. Once you
define it has to be same you cannot change it. You need to create once again if
you need to change it.
- When you convert a UNV to a UNX, all the
features and definitions (tables, joins, object definitions, security, options,
etc.) are converted. There are, however, a few limitations with the UNV to UNX
conversion:
- IDT cannot convert OLAP
universes created with prior releases. SAP recommends making a direct
connection to the OLAP source to then take advantage of the dimensional
modeling capabilities built in to IDT.
- IDT cannot convert
universes based on stored procedures of JavaBeans.
- Linked universes.
Although IDT can convert linked universes they are not supported in BI 4.
When you convert a linked UNV, IDT copies the core universe into the
converted universe. The core universe factorization is lost.
