Concepts were presented for the development of computer families, which can be used for the flexible employment by re-usable components for the derivative by computer families. In addition belonged paradigms, concepts, Tools and architectures, whose proceeding and applicability in the course of the software development process were described. Here pro and cons from the literature were summarized. In addition according to standard used concepts of the configuration management of Windows and Linux came.
For the FORE model it can not be able to be derived from the results of analysis to be derived that a consistent modelling of the requirements for fundamental architecture decisions is important in the first phases of the development process from the outset, computer families component-based there from traditional software projects.
The use of components saves thereby time and money and canalizes for knowledge of preceding program developments. Cleanly arranged systems can be waited, extended also later better and modified. Generators, which are based e.g. on GenVoca, focus thereby the complete derivative of family members and need for it consistent rule-based grammar.
Therefore different aspects and summaries should not only take place from the outset from conventional functional view, but also model other criteria such as aspects, function groups, layers or architecture basic structures, in order to facilitate themselves so the following decision for rough architecture and the technologies which can be used and to already create for the implementation a foundation-stone for the proceeding of the program design at the time of conclusion of the requirement collection. The decision for the development of computer families should already be certain in the planning phase.
The moreover one this chapter pointed out, which concepts for the deposit of software in the two usually spread operating systems are at run-time used. Generally one can assume each problem requires its specific solution here. A simple application program, which does not need any communication with other programs, is well advised with the simple and clear structure of a text file. On the other hand power requires an operating system of a high performance, like also complex structure of a data base. Also from the safety aspect it is more meaningful to deposit system-far attitudes hidden and coded. The coding cannot be used with linuxbasierenden software projects however, since various licenses forbid this also at further developed software.
Meaningfully in connection with the production of computer families the development of a catalog would be similar the efforts of the FH Mannheim, in which over certain criteria in a model the correct aids can be selected. For this the appropriate knowledge would have to be gained and prepared. This could e.g. by extended and evaluated characteristic models for listing the software requirements and system properties happened¹⁵⁹.
The concepts presented here are summarized in the following: