GeneXus philosophy, since it was born, is encapsulating complexity and automating everything that can be automated.
If we think of GeneXus transactions we see that there is a great complexity in them: the access to the data, the transactional integrity, the concurrence control, the business rules, etc. We always wanted to achieve the same with all the referent to extracting information from the databases. The first implementation to achieve this goal was the reports, and later, the workpanels, which not only allow extracting information, but also interacting with it.
In spite of the fact that we wanted to get farther, the existent base technology was not enough to develop a more advanced technology. But as the development of the base technology was getting greater (for example more powerful database managers, client/server, graphic interfaces, etc.) we realized that we were in condition to substitute the implementation of the processes for extracting information –which were done by programming infinite reports- by more advanced technology. We could develop a technology to satisfy our users in their requirements: need for accessing all the information in a dynamic and intelligent way without the need for the technicians to be involved. Besides, when working with GeneXus we ran with a great advantage, we had 'knowledge bases'; this is to say all the needed knowledge to command this technology. This made not necessary for the users to create and maintain manually complex metadata. The metadata needed for this technology could be inferred and maintained manually from the knowledge base.
With these elements and the concrete need of one of our clients the first version of this technology came up in 1997 and it was called GXOlap 1.0.
This tool had its own interface (grid) that allowed specifying a query in the user's language. The user did not need to have any technical knowledge. It guided him in the specification and generated automatically the needed sentences to satisfy the queries (in all the dbms supported by GeneXus) and its result was visualized in a grid. It also had some manipulation features such as rudimentary pivoting, averages, graphics, passing data to Excel, etc.
This tool kept on adding functionality and this is how GXOlap versions 1.1, 2.0 and 2.1 came up. In spite of being powerful in a query's specification and in its extraction (in fact almost as much as GXplorer is today) GXOlap's interface was not pretty and powerful enough in the manipulation. For that, we used the passage of the extracted data to Excel, which provided us all the manipulation beauty and power that was lacking.
When in the middle of 1998 the first Excel beta version 2000 –which added a very adequate 'dynamic table' for visualizing data, to the classical manipulating power of the electronic spreadsheets- was released, we decided to pass GXOlap's grid to Excel 2000, adding a toolbar of GXplorer, in order to command everything from Excel.
This had several advantages: firstly, because the final user is very used to using electronic spreadsheets and secondly, because these ones have an enormous data visualization and manipulation power. Finally, it allowed us investing our time in a more powerful knowledge management instead of worrying about the data visualization and manipulation.
This way, in the middle of last year, the first GXplorer (1.0) was released. It counts on all GXOlap's base technology, with its interface on Excel 2000 with its dynamic tables and to which we also added more functionality. Then, the versions 1.1, 2.0 came up. There, we powered the query's specification with better interfaces, more functionality for extracting information, more power in the security, etc.
Since the end of 1999, GXplorer has been a part of 'Microsoft Data Warehouse Alliance 2000' as a tool for building and managing DW, being the only Latin American product that integrates that alliance.
Today, version 3.0 RC is released. One of its main features is GXplorer -Web, the possibility of executing GXplorer with all its power on a navigator.