HierCube VCL and RadarCube VCL are very similar products intended for the same purposes. So, they may seem quite hard to distinguish. This article is meant to shed some light on the problem.
Both products operate with databases rather then OLAP servers. However, there is a big difference the way they work.
For HierCube there is only one fact table and it serves as the data source. The cube can also read the dimension tables but only if the fact table has all the references or the foreign keys to them. So if you want to add a dimension table into your cube, you must provide the fact table with all the foreign keys. This is called the "star " schema because all dimension tables are directly connected to the fact table.
RadarCube doesn't require direct connections between the fact table and dimension tables. Instead it supports the original structure of data warehouse (usually called the "snowflake "schema), and can create a dimension from any table related to the fact table, not necessarily directly. This makes RadarCube the natural choice when the data schema cannot be transformed to the star template.
Unlike HierCube RadarCube supports all the structure elements of enterprise OLAP servers (dimensions, hierarchies, hierarchy levels, etc.). That's why those who are used to OLAP servers like MS AS are better off with RadarCube. It is also the explanation why RadarCube's Desktop version (working directly with a database) and MS Analysis version (working as a client to an OLAP server) share the same core. This makes it easy to migrate from one product to another and even convert data from an OLAP server to a local cube and back.
HierCube doesn't support many of the entities of an OLAP server structure and can never be used together with an OLAP server.
The difference in cube structures implies the difference in capabilities of the products. Both RadarCube and HierCube support calculated measures (computed from the database records), but only RadarCube can create new calculated members based on the values of already existing ones. RadarCube allows to group measures and hierarchies as you like, which hepls you analyze your data. And finally RadarCube's Grid for OLAP data dislay is way more powerful in comparison to HierCube's.
As a result, HierCube is a low-price, but still effective way to bring OLAP functionality into your applications. Compared to RadarCube it uses a simplified internal structure and doesn't cover the whole source data in building the cube (though the snowflake schema is still possible to use). This can be viewed as considerable limitation, but on the other hand this very option allows you to create a cube simply by defining a fact table and dimension tables with standard Delphi means.
Hope this helps you make the right decision. Anyway, if you have any questions, feel free to contact us at sales@radar-soft.com.