RESEARCH INTO THE METHODS OF SOFTWARE PRODUCT DEVELOPING AND MAINTAINING

Texto integral

Introduction. Information technologies are one of most dynamic developing areas of science, technology and engineering, besides they are included in the list of critical technologies, which contributes to the improve- ment and continuity of new solutions, aimed at creating advanced information systems [1]. Therefore, the applica- tion of an integration environment of information coop- eration (IEIC) based on Open Source technology in the aerospace segment is an actual solution that will solve the issues of information support in managing projects of various complexity and direction, regardless of the subject’s location. IEIC system is the cheapest one and has all necessary functions and capabilities, taking into ac- count tight budget and human resource constraints. In this respect, in order to form the information environment of the IEIC it is necessary to investigate the basic strategies and their models for the process of developing and putting into operation a software product. This will enable to choose the most effective design model of the IEIC sys- tem development, to define the design environment and to describe the project characteristics, the concept and the tasks of implementing the IEIC prototype [2]. Analysis of the efficient method selection for the IEIC system creation and development. According to the Russian software development stan- dards, there are three basic strategies: cascading, incre- mental and evolutionary models [3]. The cascading strat- egy is a linear sequential design approach for software development; the incremental strategy assumes that the requirements for the software product are implemented gradually, each time expanding a product utility; in the evolutionary strategy, the requirements are not fully de- fined, but are dynamically refined during the development of software product versions [4; 5]. Figure shows basic models of the basic software development strategy. However, although there are many different models and ways of software products development, there is no single method that describes the Web resource develop- ment model using Open Source technologies, such as De- velopment of IEIC. Therefore, to select a life cycle model Software Quality Institute in USA recommends to classify the project and to identify the main parameters that will be the key factor to select the model of software develop- ment [6]. These include: 1. Requirement parameters for the project. This cate- gory can include the possibility to set up requirements for software at the beginning of the software life cycle, to determine the probability of changes and additions to the requirements for the software development lifecycle. 2. Parameters of the project working team. This category determines the novelty of technologies and the novelty of toolbox development for developers. 3. User profiles. This category identifies the degree of users involvement in the development process and their relationship to the project team, determines the degree of user involvement in the process of developing software and acquaintance of users with the problems in the proc- ess of the software life cycle. 4. Parameters of project types and risks. This category reflects the complexity of the project, estimates the re- sources for its execution, identifies problems in the soft- ware domain and determines requirements for reliability levels and other criteria [7]. Based on the above said a summary table that includes all categories of project parameters is suggested. Table will help to choose suitable model of software lifecycle for a particular project [8-10]. Software development models Models of software engineering № Classification parameters Model development of life cycle of the software product Cascading V-model Iterative Incremental Spiral RAD Evolutionary 1 Are requirements for the project defined and feasible? Yes Yes Yes Yes No Yes No 2 Can requirements for the project be determined at the beginning of the software life cycle? Yes Yes Yes Yes No Yes No 3 Do you need to demonstrate requirements for clear understanding? No No No No Yes Yes Yes 4 Is it necessary to check the concept of the software product? No No No No Yes Yes Yes 5 Can the requirements be changed or added in the process of the software life cycle? No No Yes Yes Yes No Yes End table № Classification parameters Model development of life cycle of the software product Cascading V-model Iterative Incremental Spiral RAD Evolutionary 6 Is there a need to implement requirements in the early stages of software development? No No No Yes Yes Yes Yes 7 Are the tasks of the subject area of the project new to the project team? No No Yes No Yes No Yes 8 Are the tools for the implementation of the project new to the project team? No No Yes No Yes No No 9 Can participants change roles in the process of software lifecycle development project? No No Yes Yes Yes No Yes 10 Is in the process of the software lifecycle evolution and validation stage of software develop- ment needed? Yes Yes No Yes Yes No Yes 11 Will users participate in the process lifecycle? No No Yes Yes Yes No Yes 12 Will users evaluate the current status of the software in the process of software development? No No Yes Yes Yes No Yes 13 Will users be involved in all phases of the software life cycle development? No No Yes No Yes Yes No 14 Will users track the progress of the project? No No Yes No Yes No Yes 15 Is the software being developed new to the company? No No Yes Yes Yes No Yes 16 Will the project be an extension of the existing system? Yes Yes No Yes Yes Yes No 17 Will the project be large-scale? No No Yes Yes Yes No Yes 18 Will the project be medium or small-scale? Yes Yes No No No Yes Yes 19 Will the product life cycle be long-lasting? Yes Yes Yes Yes Yes No Yes 20 Is a high level of product reliability necessary? No Yes Yes Yes Yes No Yes 21 Is it planned to upgrade and develop the product during the operation phase? No No No Yes No No Yes 22 Is the timetable tight? No No No Yes No Yes Yes 23 Will the functions and modules be reused? Yes No No Yes No Yes Yes 24 Is the project budget tight? No No No No No No Yes 25 Are the software developers sufficiently competent? Yes Yes Yes Yes Yes Yes Yes 26 Can costs be required to purchase additional equipment and tools for project implementation? No No Yes Yes Yes Yes Yes 27 Can additional costs be required to attract highly qualified employees? No No Yes No Yes Yes Yes 28 Have such software products been implemented in the enterprise before? Yes No No No No Yes No 29 Will new technologies or new approach be used in the development process? No Yes Yes Yes Yes No Yes 30 Is it necessary to conduct analysis of existing technologies for software development? No No Yes Yes Yes No Yes 31 Is it necessary to develop a prototype of the software product in the early stages of the product life cycle? No Yes No Yes Yes Yes Yes 32 Is the software unique? No No Yes No Yes No Yes 33 Is the software part of the system? Yes Yes No No Yes No No 34 Is the software a stand-alone solution? No No Yes Yes No Yes Yes According to table the cascade model is more suitable when the requirements are known, straightforward and documented, and there are no problems with the availabil- ity of programmers of necessary qualifications [11-13]. This model can be used in relatively small projects, in the redevelopment of typical software (for example, the de- velopment of an electronic document management, ac- counting system, etc.). The cascading model can be used in case of migrating of existing software to a new plat- form or when launching a new version of software if the changes are minor and clearly defined. The advantage of this model is that each stage is completed with certain documentation that meets the criteria of completeness and consistency. Therefore, there is no need to return to the previous stages. All work is performed in strict order. This approach allows to schedule deadlines, and accord- ingly plan the costs of each stage in advance. The disad- vantage of the model is the complexity of switching be- tween stages [14]. The V-model fits better when requirements are de- fined and documented, thorough testing is required. The model is suitable for medium and small-scale projects, 708 provided that qualified programmers, including testers are involved. The advantages of this model include require- ments setting prior to development, tests planning and system verification at the early stages of software devel- opment, special importance is given to time management. However, there are disadvantages: the model does not imply the work with parallel events; there is no possibility to introduce requirements for dynamic changes at differ- ent stages of the software life cycle. The model can be used mainly in projects with tight time or financial con- straints, also in such projects where there is an extensive coverage of the test tasks. The iterative model can be used for large-scale pro- jects, when the requirements for the final system are clearly defined and understood, but can be modified over time. That is, the model implies that the main task must be defined, and implementation details may change during the development of the software product over time. The incremental model is the most suitable when basic requirements for the system are clearly defined and if the software needs early introduction on the market, however, there is a possibility of revision over time. This model can be used in long-term projects at low or medium risks. Also, the model is suitable for projects where new tech- nology can be applied. This will allow the user to adapt to the software product being developed by executing smaller incremental steps at specific time intervals. The advantage of this model is that the result of each incre- ment is the working functional product and a customer has the opportunity to evaluate each developed version of the software. Firstly, it is supposed to develop and im- plement the main function of the platform in the model and only then increase it. This will reduce the risks of failures and risks of over-expenditures, as the main func- tionality of the system can be determined at the early stages of development. The disadvantages of the model include the fact that at the initial stage it can be difficult to define a complete efficient system which makes it diffi- cult to make up the increments. The spiral model is suitable for complex and expen- sive projects, where it is required to carry out analysis for impact evaluation after each step. If software develop- ment involves the use of new technology and success is not guaranteed, then the spiral model should also be used. This model is appropriate for development of new series of software products, where it is important to analyze the risks and costs. In the model it is possible to return to the left behind stages to reflect over the changes in costs of risks. The advantages of the model include step by step specification requirements and conducting a risk analysis of the project, which allow to identify design errors at the early stages of software development, creation of working prototypes at the early stages and availability of com- pleted software development process documentation, as well as the ability to add new features even at the later stages of the software life cycle. A complex project can be developed in parts, highlighting the most important requirements at different stages. The disadvantages of the spiral model include the dependence of success of the development process on the stage of risk analysis and high cost of software development, because risk management requires the involvement of sufficiently qualified specialists. The RAD model can be used only in the presence of highly qualified specialists as this model is suitable for confident knowledge of the target business and the need for urgent production of system [15]. This model is ap- propriate for projects where the budget is large, in order to pay for professional service. Also, this model can be used in projects where there is very little risk; the soft- ware product being developed can be modeled and has relatively low performance; and when it is required to minimize duration of software product development. The advantages of this model include reduction of the cycle time and the number of developers by using powerful tools. Already in the early stages of software development a prototype of product, which further increases the effi- ciency of the developed components, is created. The dis- advantages include rigid time management for software development, and the need to attract highly qualified pro- fessionals who are able to work with the necessary soft- ware tools. In the evolutionary model, requirements for the soft- ware product can be specified gradually. The main em- phasis is put on the development of the software proto- type, then a complete understanding of product require- ments. The advantages of this model are: - identification of software utility at the initial stage; - staff recruitment on demand; - dividing of system into incremental components. The disadvantages of this model are limited opportunities for long-term resource mobilization [3; 9]. Conclusion. Analysis the best model selection deter- mined the evolutionary model as the most suitable one for the IEIC development. Advantages of the evolutionary model include the fact that the product is developed in the form of separate designs, but unlike the incremental model, requirements cannot be initially determined com- pletely. In this model, specification of requirements is allowed partially and is specified with each subsequent design. Since the specificity of the development of the IEIC platform lies in the fact that at the initial stage there are no exact formulations for the platform requirements, so accordingly, there are no specific tasks for the devel- opment of the IEIC, and there is no unified design for commissioning of similar systems developed with the help of the Open source technology. In this regard based on the processes of the evolutionary model, it is necessary to work out a methodology for constructing the IEIC, taking into account the specifics of the web resource and free software products development, and for the existing technologies analysis, their functions and characteristics for developing and maintaining the IEIC.

Sobre autores

Zh. Abenova

JSC “National Company “Kazakhstan Gharysh Sapary”

National Space Centre, 89, Turan Av., Astana city, 010000, The Republic of Kazakhstan

M. Petrov

Reshetnev Siberian State University of Science and Technology

31, Krasnoyarsky Rabochy Av., Krasnoyarsk, 660037, Russian Federation

Bibliografia

Arquivos suplementares