Structures as information storage mediums

We would probably drown in the increasing flood of information which we are exposed to on a daily basis if we were not able to evaluate and then reject, put aside, or arrange information. A similar thing applies to the chaos on our hard disk, which we can fight by sorting and structuring files. A tool which the operating system makes available to us is folders . We give these folders meaningful names which describe their contents, and sort the files accordingly. Subfolders relate closer and closer to a specific topic or task. Working with files is an established concept.

An information system such as would be unthinkable without the ability to administer structures and rights of access. Well thought-out structures tell us something about the documents contained therein and are outstandingly suited for passing on information, i.e. for inheriting from a folder to its subfolders (and so on). However, with you do not structure with folders in the conventional sense. Strictly speaking there are no folders, but simply data types , which you define with XML schema . The special thing in this is however: Document types can define which other document types may be located under them . This is one of the further, central characteristics of the information system.

The example of an address administration should clarify this concept

Imagine you are building up a small address administration. You define a document type »company«, which allows the entry of central customer data. Moreover, you define a »branch« type and a »department« type. With a few clicks of the mouse you define that only branches may be created below a company and below these only departments. Below a department on the other hand, further departments and also documents of the »employee« type may be created. You can surely imagine how quickly you can produce complex information structures of any type in this way.

And it all goes a step further: Imagine you are creating a project administration in another place in the system. You can define that you reference documents of the »contact« type as being involved in the appropriate project documents. Although perhaps only the name and email address are indicated in the contact used, its position in the structure tells us something about the department, the branch and the company in which the contact works. On the other hand, the system tells you which projects the contact is involved in upon viewing the contact.


If a few suitable pictograms are then integrated for all document types, an individual system for the administration of contact information and projects is created in a few minutes . And this system partakes in all advantages offered by the system:

  • Version log with view of each modification,
  • Rights administration through to object level,
  • Referential integrity and
  • Object locks.

If an application is to be created you may already be halfway there.

By the way: If this all scares you off and you would much rather work with folders, then this is also possible of course. You can create data types which can contain no data themselves, but merely define which types you can create underneath. You therefore have documents without any content, which contain further documents– that is to say folders.