当前位置: X-MOL 学术Inf. Softw. Technol. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Exploring the software repositories of embedded systems: An industrial experience
Information and Software Technology ( IF 3.9 ) Pub Date : 2020-11-15 , DOI: 10.1016/j.infsof.2020.106489
Jakub Polaczek , Janusz Sosnowski

Context

Tracing reports for software repositories have attracted many researchers. Most of them have focused on defect analysis and development processes in relation to open source programs. There exists a gap between open source and industrial software projects, which, in particular, relates to different schemes for creating software repositories and development schemes. This is especially true for embedded systems that gain large markets and become more complex.

Objective

The aim is to explore the software repositories of industrial embedded systems and derive characteristic features in order to evaluate quality and identify problems to do with development processes.

Method

In this paper we have proposed a novel approach to software repository analysis based on the fine grained exploration of issue tracking and code control repositories. In particular, we distinguish the various activities of project actors (e.g. creating new functions, correcting defects, improving performance, modifying tests) and analyse them in a context, not only of a single project, but also a set of correlated projects that have been developed in the company. These issues have been neglected in the literature. These analyses needed new holistic schemes for repository exploration, including various statistical metrics, text mining, and machine learning techniques.

Results

In exploring selected industrial projects we have identified that only 40–75% of issues relate to defects; the issue reports and commit descriptions included here comprise a lot of data that has been disregarded in the literature. These data allow us to trace diverse types of code changes and identify imperfections in software repositories.

Conclusion

We show that fine grained repository analysis gives a broader and more complete view of project development, which may lead to its improvement.



中文翻译:

探索嵌入式系统的软件存储库:行业经验

语境

软件存储库的跟踪报告吸引了许多研究人员。他们中的大多数人都专注于与开源程序有关的缺陷分析和开发过程。开源和工业软件项目之间存在差距,尤其是与用于创建软件存储库和开发方案的不同方案有关。对于拥有庞大市场并变得更加复杂的嵌入式系统而言,尤其如此。

目的

目的是探索工业嵌入式系统的软件存储库并获得特征,以便评估质量并确定与开发过程有关的问题。

方法

在本文中,我们基于对问题跟踪和代码控制存储库的细粒度探索,提出了一种新颖的软件存储库分析方法。特别是,我们区分项目参与者的各种活动(例如,创建新功能,纠正缺陷,改善性能,修改测试),并在上下文中对其进行分析,不仅是单个项目,还包括一组相关项目。在公司开发。这些问题在文献中已被忽略。这些分析需要新的整体方案来进行存储库探索,包括各种统计指标,文本挖掘和机器学习技术。

结果

在探索某些工业项目时,我们发现只有40-75%的问题与缺陷有关。此处包含的问题报告和提交说明包含许多在文献中已忽略的数据。这些数据使我们能够跟踪各种类型的代码更改,并确定软件存储库中的缺陷。

结论

我们表明,细粒度的存储库分析可以为项目开发提供更广泛,更完整的视图,这可能会导致其改进。

更新日期:2020-11-23
down
wechat
bug