Situation

With maintenance of existing program systems, questions often arise about the effect that any specific changes have on other components. The amount of effort to answer such critical questions, indeed whether they can be answered or not, depends on the quality of the existing program documentation.

Typical questions during a change process are:

A change has to be made to a

	copy book:	Which programs use that copy-book?
	file:	In which programs is that specific file used?
	sub-program:	In which programs is that specific sub-program used?
	screen mask:	In which programs is that specific screen-mask used?

Which other programs (sub-programs, copy-books, screen-masks etc.) are affected by any of these changes?

Which other programs (sub-programs, copy-books, screen-masks, etc.), which are affected by any of these changes, are processed by other users?

Which other programs (sub-programs, copy-books, screen-masks, etc.) have to be recompiled and/or re-cataloged after such a change is completed?

Which other configurations (projects, documentation) are affected by such a change?

Missing information makes maintenance extremely difficult and very costly, increases time spent on further development and makes the processing of other requirements standard.

This is an unnecessary security risk!

Solution using SET-LIBER Online-Repository

The Repository information (eg. Cross-References, Processing Statuses, History) for documents managed by SET-LIBER is maintained automatically whilst working on these documents. Transparency of the co-relationships of in-house and/or third-party software elements, dramatically reduces the source of errors. The effort required for maintenance and documentation is substantially reduced, this will reduce the cost! Groups of documents managed within SET-LIBER are always differentiated by freely definable Sourcetypes. For analysis purposes by the Repository the user can also select one related Document Type (Docu Type) for each source type. Examples of Document Types are:

Program Copy Book File Screen Mask Transaction SQL Table Computer Center Documentation Job Control Stream Job Step Job Procedure Project Description Configuration Description

The Online-Repository which is integrated in SET-LIBER has the following functions and benefits:

The Cross-References are always up-to-date and complete (including version and processing statuses of all dependent documents).

Document sources drive the automatic documentation (Repository).

The documentation system has a common management system for program sources and Job Control Streams.

The creation and maintenance of the end-user documentation is controlled and supplemented, automatically by computerised steps (eg. by Cross-References to related configuration and documentation descriptions).


Manual Documentation

Documentation and Configuration Descriptions which are created and maintained by the end-user, can be stored in text files. The end-user decides which information is stored in these text files.

The SET-LIBER concept recommends that, in text files, the demand descriptions of the requests are stored. In this case a specific input format (screen mask) for these text files can be defined by the SET-LIBER administrator. With each process, the predefined mask will be used.

Documentation descriptions are referenced automatically. The automatic analysis of documentation texts makes the temporary and/or permanent project connections.


String Search Function

A question often asked is, in which programs (copy books etc.) is information with a specific name (character string) used. The String Search function within SET-LIBER provides the end-user not only with a way to answer these questions but with a function provides for targetted search results, using the Repository.


Summary

The Repository information is always up-to-date and complete - this is an automatic benefit.

The creation of the documentation information from the source data is completed without any manual effort. When changing specific documents it is immediately apparent which documents have been changed, and which related documents are affected.

The automatically recorded information about the time of the last change to a specific document, makes any inconsistencies to any and all dependent documents immediately apparent.

By saving the history of each document, it is always very clear which version of a program, sub-program, copy-book, etc. was effective, and when. By using the "input text files" (Docu Texts), information may be added to the documentation, when it is not clear from the source code.

The String Search function permits a string search down to the statement level.

The Repository offers complete overviews (catalogues) of all documents, these are subject to selection and can be sorted.