We've heard of Agile methodologies for several years, but software projects have still delays and failures in planning and management.How then automate the processes of the development of software projects following agile methodologies?In this paper we try to answer this question by proposing an appropriate use of collaborative environments for software development that effectively translated into action the principles and practices of Agile methodologies. The appropriate use of these collaborative environments leads to a truly agile management, where the project manager achieves the vision and the continuous monitoring of the project, and the participants become adapted to an agile and productive discipline imposed by the environment, regardless if they are geographically dispersed. To ensure that appropriate use, this paper proposes a methodological guide to achieve an agile management of software projects by relying on a particular collaborative environment.
This paper presents the practical experience and results of the Lion Project, which aimed to improve software development times at Heinsohn Business Technology (HBT), a large-scale Colombian software development company. The main result of this project is the LionWizard Framework, a set of libraries and tools with a focus on large-scale software reuse and integration. The Lion- Wizard Framework integrates all of the existing libraries at HBT using Maven and provides a Wizard. The latter uses code generation and program transformation to automatically integrate all of the required components into an initial codebase in Java EE. The Wizard provides sufficient exibility to seamlessly integrate future components into the development process. Before the Lion Project, software developers required several days of even weeks to integrate all of the required components into a code base for each project. This new framework reduces those times to a few hours.
Purpose
This purpose of this paper is to present ISML-MDE, a model-driven environment that includes ISML, a platform-independent modeling language for enterprise applications; ISML-GEN, a code generation framework to automatically generate code from models; and LionWizard, a tool to automatically integrate different components into a unified codebase.
Design/methodology/approach
The development comprises five stages: standardizing architecture; refactoring and adapting existing components; automating their integration; developing a modeling language; and creating code generators. After development, model-to-code ratios in ISML-MDE are measured for different applications.
Findings
The average model-to-code ratio is approximately 1:4.6 when using the code generators from arbitrary models. If a model transformation is performed previously to the code generation, this ratio raises to 1:115. The current validation efforts show that ISML properly supports several DSL essential characteristics described by Kahraman and Bilgen (2015).
Research limitations/implications
ISML-MDE was tested on relatively small applications. Further validation of the approach requires measurement of development times and their comparison with previous similar projects, to determine the gains in productivity.
Originality/value
The value of ISML-MDE can be summarized as follows: ISML-MDE has the potential to significantly reduce development times, because of an adequate use of models and transformations. The design of ISML-MDE addresses real-world development requirements, obtained from a tight interaction between the researchers and the software development company. The underlying process has been thoroughly documented and it is believed it can be used as a reference for future developments of MDE tools under similar conditions.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.