Previous research of software being developed under open source license has showed that some open source products can reach qualitative and popularly level of commercial (usually closed-source) counterparts. Open source (usually free of charge) software has become increasingly prominent in last few years. However, a lot of open source software does not attain maturity. This shows that their projects have problem and that the quality of projects outcomes (software) can vary. Therefore an assessment model is necessary which could filter out only suitable (qualitative) products. This paper presents a quantitative basis for evaluating an open source product based on multiple criteria of software quality. In contrary to other software assessment models, which are time-consuming and expensive, the proposed model uses easy accessible quantitative data, specific for open source projects. A set of measurements for evaluating open source software together with proposed model and a practical example is presented.
Keywords: Open source software, software quality, quality assessment model
[...] Using the Internet, this peer-review process harnesses thousands of highly trained individuals to test and review all existing and new ‘blueprints' to insure a very high level of security, quality and reliability Qualitative aspect of OS software development Public available source code enables new and innovative approach to software development, which might be in some aspects even better than “traditional” approaches. As summarized in Eric S. Raymond's work Cathedral and theBazaar” (CatB) some positive aspects of OS software development model (associated with the Bazaar model) are following OS software development model is based on flexible approaches, which enable frequent releases and simplified delegation of tasks. [...]
[...] Objectivity; the evaluation result is factual (single scalar value), presenting most preferred alternative Application of assessment model An application of the proposed model involved a comparative assessment of several open source web application frameworks. The objective of the assessment was to select the best framework using proposed model. For demonstration, seven frameworks were selected, based on Java web technology Despite of higher popularity of some frameworks, we did not preferred any alternatives. The selected alternatives (in alphabetical order) were following: a. [...]
[...] Further, while the goal of the quality is always to achieve end-user needs, software users quality requirements should play a major role in quality assessment model. This aspect of software quality is evaluated by quality-in-use. Ouality-in-use (see also Figure 2 presentsuser's view of quality and might vary among different users. Quality-inuse metrics measure the extent, to which a product meets the needs of specified users to achieve specified goals with effectiveness, productivity, safety and satisfaction in a specific context of use [lt]. [...]
[...] Assessment model should be additionally adapted to existing quality standards Conclusion We presented a model for comparative assessment of open source products. Despite of other software assessment models, which are time-consuming and expensive the proposed model uses simple accessible quantitative data specific for OS projects. The model is based on quantitative data and consists of 6 steps. The outcome of the model, which presents the most preferred alternative, is calculated in two main phases. First software quality characteristic values are obtained from identified OS metrics. [...]
[...] Since these types of attributes can be easily gathered for OS software, they can simplify Specification of the proposed OS assessment model The proposed OS assessment model, is made up of six phases, which are performed sequentially [Figure 31: Figure Proposed OS assessment model 1. The objective of the first phase is to identify possible OS software altematives (AI). The alternatives are usually identified using open source portals and www search engines. The phase acts as a pre-assessment phase where alternatives, which do not conform to functional and nonfunctional requirements, are eliminated (e.g.: the altemative does not support specified operation system, programming language, maturity status or user functions) Quality metrics (Mja)re identified in second phase. [...]
using our reader.