All is not lost however, although hard times are ahead. Reporting Services is tied into SQL Server, and while there are a lot of users of that database software, it is not the dominant one; far from it. And in the future there is a major threat to SQL Server itself in the form of MySQL running on a Linux platform. For the time being then there is still a market for Cognos, etc. with Oracle, DB2, MySQL and the older RDBMS products such as SyBase. However they must reconcile themselves to the loss of the Microsoft market.
Even though Reporting Services is free, users can't throw out the other products overnight, because they have developed applications which will take some effort to port. Besides we don't yet know if the new product is any good.
Microsoft has already made a move against the vendors of more sophisticated tools, referred to as On Line Analytical Processing (OLAP), by binding a layer of analytical functions into SQL Server. This however is in common with Oracle and IBM who have similar extensions to the core SQL databases to support time series analysis, etc. But these are aids to OLAP, similar to the Business APIs added to application packages such as SAP; they are not direct competitors to full function OLAP tools. In any case these more advanced tools work either with data warehouses or across multiple databases. As such they are not as vulnerable to the Microsoft competition as are the advanced reporting tools such as Business Objects, etc. There is a twist in this though because the most common report writing tool, Crystal Reports, was used extensively by Microsoft, Crystal Decisions, a more advanced tool, has been taken over by Business Objects.
Reporting tools are particularly useful for generating formal reports from transaction databases. But they are not analytical tools. The transaction databases update data, while an information database needs historic data in a recognisable form, which is the role of the data warehouse, not a transaction SQL database, even those with the so-called OLAP extensions. The differences lie in the data modelling more than in the database management system itself. The dream of a common transaction and analytical database is a long way off yet. Thus there remains the need to move data from one or more sources to populate the data warehouse. This is a non-trivial task! The source data needs extracting, cleansing, translating and loading (ETL). Many organisations wrote COBOL programs to perform these functions, but that does not cater for the complexity and multiplicity of data sources in today's world. There have been a number of excellent tools developed to perform ETL, but they have suffered from the same old problem, they are far too expensive! It is therefore not surprising that Microsoft is readying ETL products for SQL Server. These functions will probably be packaged, along with Reporting Services, with the next release of SQL Server, code named Yukon. These are all good moves by Microsoft.< BR>
Martin Healey, pioneer development Intel-based computers en c/s-architecture. Director of a number of IT specialist companies and an Emeritus Professor of the University of Wales.