Skip to main content
European Commission logo print header
Content archived on 2024-04-15

Replay and Evaluation of Software Development Plans using Higher-Order Meta Systems

Objective

The objective of REPLAY was to provide evidence of the feasibility of reusing development plans and their component modules within the process of creating industrial software. The project aimed to advance the rather poorly understood domain of reusability.
Beside building on progressive generalisation of case studies, REPLAY explored both top-down replay of developments and bottom-up assembly of components. The possibilities of continually controlling these development plans was investigated by means of abstract interpretation and modelling of operational properties.
This project made extensive use of the DEVA language defined in the TOOL-USE project (510).
The objective was to provide evidence of the feasibility of reusing development plans and their component modules within the process of creating industrial software. Current domains under analysis are replays of development through several different technical approaches: the transformational style of developments; proof steps for data type representations or rectification in the Vienna development method (VDM) context; and composition of programs from components through a graphical connection language. The examples used for the observation and experimentation of these techniques are different development scenarios of a common problem extracted from a large application in biology; human leucocyte antigen (HLA) typing. This problem has been analysed using a symbolic approach with a different implementation target. This case study has been used to show the integration of techniques of the project, namely: transformationdevelopments driven by propogation of operational properties; an integrated top down and bottom up style of development using functional and relational styles for the composition of development parts; integration of methods (VDM algebraic methods) and support of these methods. A set of tools was designed and implemented for the support of the various approaches: the DEVA environment has been modified to support specific operations for reusing formal developments; the LPG environment has been modified and extended by an information system to support the reuse of algebraic components; and a library interconnection language (LIL) has been implemented, offering a graphical interface with facilities for the bottom up relational assembly of components.
Current domains under analysis are replays of development through several different technical approaches:
-transformational style of developments
-proof steps for data type representations or rectification in the VDM context
-composition of programs from components through a graphical connection language.
The examples used for the observation and experimentation of these techniques are different development scenarios of a common problem extracted from a large application in biology: Human Leucocyte Antigen (HLA) typing. It is presently tackled on large computers by statistical techniques. This problem has been analysed in the REPLAY context using a symbolic approach with a different implementation target.
This case-study has been used to show the integration of techniques of the project, namely:
-transformation developments driven by propagation of operational properties
-integrated top-down and bottom-up style of development using functional and relational styles for the composition of development parts
-integration of methods (VDM algebraic methods) and support of these methods.
During the final period of the project, a set of tools was designed and implemented for the support of the various approaches:
-the DEVA environments, coming from the companion project TOOL-USE, has been modified to support specific operations for reusing formal developments
-the LPG environment, designed by LIFIA, has been modified and extended by an information system (F1) to support the reuse of algebraic components
-a Library Interconnection Language (LIL) has been implemented, offering a graphical interface with facilities for the bottom-up relational assembly of components.
Exploitation
Although no industrial product has stemmed from the project, the knowledge gained should be valuable in achieving reusability in the longer-term and brings nearer the dramatic improvements of productivity and quality that could follow.
The project also helped to start the industrialisation of the software related to the HLA case study by a new project launched in the SPRINT research programme.

Topic(s)

Data not available

Call for proposal

Data not available

Funding Scheme

Data not available

Coordinator

CISI Ingénierie SA
EU contribution
No data
Address
13 rue Villet
31029 Toulouse
France

See on map

Total cost
No data

Participants (4)