基于价值的软件测试管理外文文献翻译中英

基于价值的软件测试管理外文文献翻译

(含：英文原文及中文译文)

文献出处： Ramler R, Biffl S, Grünbacher P. Value-Based Management of Software Testing[M]// Value-Based Software Engineering. Springer Berlin Heidelberg, 2006:225-244.

英文原文

Value-Based Management of Software Testing

R Ramler ， S Biffl ， P Grünbacher

Abstract

Testing is one of the most resource-intensive activities in software development and consumes between 30 and 50% of total development costs according to many studies. Testing is however often not organize the maximization of business value, nor is it to shoulder the mission of the project. Path testing, branch testing, guidance testing, transformation testing, scenario testing, and requirements testing are all equally important in all aspects of software. However, 80 percent of the value in practice often comes from 20 percent of software. In order to get the maximum return on investment from software testing, test management needs to maximize its value contribution. In

1文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

this chapter, we will further promote the need for value-based testing, describe the practices that support value-based test management, outline the framework for value-based test management, and illustrate the framework.

Keywords: value-based software testing, value-based testing, test costs, test benefits, test management

Introduction

Testing is the most important and widely used method in the software quality assurance process. The checksum verification aims to ensure that the software runs correctly and ensures the quality of the software and the reliability of the software through comprehensive analysis. In IEEE 610.12 (1990), a test is defined as the activity of observing and recording an executed system or component under specified conditions and evaluating the system or component.

Tests are widely used in practice and play an important role in many organizations that ensure quality strategies. Software affects the daily lives of tens of thousands of people and shoulders arduous tasks. Therefore, software will be particularly important in the near future. Research shows that testing usually consumes 30% to 50% of software development costs. Even higher proportions of safety critical systems are not surprising. Therefore, the challenge of software testing is to find more effective ways to conduct effective testing.

The value of software test management is to strive to reduce test costs and

2文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

meet demand. Valuable test management can also provide good guidance on project goals and business value. In the first chapter, Boehm cited many potential testing costs. This example illustrates the use of 7% of the customer checkout type to complete a 50% software test benefit. Although 100% testing is an impractical goal, there is still room for improvement and savings to achieve the desired value by adjusting test methods. The value power based on software engineering lies in the fact that current software engineering practice research is equally important in terms of requirements, test cases, test objects, and product defects. This pair of tests is clearly correct and it contributes indirectly to the value of the product. The separation of development and testing makes this problem even more prominent. Testing is often a purely technical issue that will make the relationship between testing and business decision-making more closely linked. This chapter mainly describes the need to improve value-based test management, explains its basic elements, discusses existing examples to support value-based testing, and the basic framework of value-based test management. The basic structure of the rest of this chapter is as follows: Section 11.2 discusses the value of test management contributions; Section 11.3 discusses the case studies of test management support; Section 11.4 describes the value-based test management framework with examples. At the end of this chapter, we will further study the specifics.

Description Based on Value Test

The goal of value-based checksum verification is defined as ensuring that the implementation of the software meets its intended target value benefits. If we consider from the perspective of value, then where should the contribution of the

3文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

test be? Basically, it can be considered from two aspects: The internal aspect includes the test cost and the test benefit, the external aspect emphasizes the future system threat and the opportunity. The key to value-based testing is to combine these two aspects, that is, to adjust the internal testing process through customer and market requirements.

In order to combine both internal and external factors, it is obviously not appropriate to focus solely on technical tests. On the contrary, test management needs a global grasp.

Example 41 describes the dependencies between the internal and external aspects of test management. The internal aspect is like the test manager's control over the entire project. The expenses in this area mainly come from software testing practice and short-term and long-term testing. The external aspect is considered to be some of the benefits and parameter data that the test supervisor can control. Value-based software test management enables testing to satisfy value-based value propositions and focus the entire team on valuable test directions.

The first question for an external view of software testing is “How do we ensure the value objectives of the software system?” This goal is to test the valuable parts of the software focus, the most important qualities, and the timely adjustment of project risks by coordinating value propositions. Answering such questions includes market opportunities, project value propositions, and cost-effectiveness. Refer to the first chapter for a detailed introduction to

4文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

opportunities and risks, and refer to Chapter 7 for the introduction and reconciliation of value propositions. Internal views are based on the benefits of value propositions, and the test budget represents an approximate level of the entire project. The main problem is how to use testing as an investment activity. In order to be able to test efficiently and quickly and reduce the development budget. Appropriate internal and external communication coordination can satisfy the benefits of testing.

The value contribution of the test

Compared to other development processes such as code and user interface design, testing does not immediately create value for the product. However, testing provides and supports valuable task activities generated during software development. The key to understanding the value of test contributions is the contribution of the test. The contribution of the test establishes the relationship between the test and the value of the final product value.

The most direct customers are program developers and project managers who are directly linked to the test team. The concentration of power in value-based software engineering testing is the customer and the user (see Chapter 7). Customers and users conduct testing by setting the context and scope to reach the value goal of the test.

Tested customers

5文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

Developers, project managers, quality supervisors, customers, analysts, end-users or maintenance staff all benefit from the analysis of software systems, rely on feedback to detect problems, reduce uncertainty, and make relevant decisions to accelerate product development. .

The following example shows feedback from different groups for test requirements:

The extent to which the customer and the user are satisfied with the need for consistency and to what extent the software's value benefits are satisfied. Testing also provides visibility and insight into the progress of the project. The test results can be used to understand the test cases that have passed. When the acceptance test appears unpractical or fails to appear in the actual environment, the α and β tests provide a more solid basis to verify the results. ∙ Sales and product executives obtain relevant information from test plans, pricing, promotions, and distribution. The difference between the actual quality of the product and the quality expected by the customer and the user can easily lead to misunderstandings and misconceptions that reduce or prevent the realization of true value. In order to be able to successfully meet these expectations and meet the goals of individuals or organizations, product design is adjusted to customer needs to meet certain functions. For project managers, testing supports risk management and project process estimation. The focus is on identifying and eliminating the risk of potential damage and the suppression of value merit. The serious shortcomings that significantly reduced project performance in the early stages were a major goal. Testing reduces uncertainty and helps project managers make better, smarter

6文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

decisions about removing defects, system stability, and product updates.

Quality managers are more interested in the identification of issues and dynamic trends in specific issues. The test results helped the evaluation of the project, the assurance of the quality strategy and the improvement of the process. Rosenberg discusses how testing can contribute to ensuring quality and shows how test problems can be verified as how to improve the process of the project. Developers and users understand the current state of related issues and provide relevant data to measure and predict software quality and reliability.

Developers often need to obtain feedback to verify that the test implementation is complete, meets standards, and meets quality requirements. In order to ensure stability, the test provides detailed information about the defect and indicates why the test failed. In addition, tests provide feedback on the improvement of project defects. For example, it is necessary to pay attention to the need to test whether the original function has been changed or the decline has occurred with the relevant changes.

For demand engineers, testing is valuable for verifying and confirming requirements. Weinberg once pointed out that \"one of the most effective methods is to pass test cases just like testing a complete system to form test requirements.\" Black box testing can help ensure their integrity, accuracy, transparency, and conciseness through the help of their needs. So testing can increase requirements and move toward test-driven.

7文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

In short, testing can improve benefits by making plans less uncertain and risky, make related decisions, and strive to control the minimization of unnecessary consumption (internal causes). More importantly, it helps to realize the expected value benefits. The benefits of these benefits are not free, and the cost of testing is often significant (external reasons). Tests can be understood as buying information and can also be considered as reducing cost risks and reducing uncertainty of investment activity. Investment in costs and benefits requires making relevant decisions about test needs. So the next two questions are: What is the cost of testing and what is the test benefit of value activity?

中文译文

基于价值的软件测试管理

作者：R Ramler ， S Biffl ， P Grünbacher

摘要

根据研究表明测试已经成为软件开发过程中一个很重要的环节, 它占据了整个软件开发成本的百分之三十到五十。 测试通常不是用来组织商业价值的最大化, 也不是肩负着项目的使命。路径测试、分支测试、指导测试、变换测试、场景测试以及需 求测试等对于软件的所有方面都是同等重要的。 然而在实践中百分之八十的价值往往来自百分之二十的软件。 为了从软件测试中得到最大的投资回报, 测试管理需要最大化它的价值贡献。 在本章, 我们将更加促进对基于价值的测试的需要, 描述支持基于价值的测试管理的实践,勾画出基于价值的测试管理的框架,并举例说明该框架。 关键词 : 基于价值的软件测试 , 基于

8文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

价值的测试 , 测试成本 , 测试利益 , 测试管理引言

测试是软件质量保证过程中最重要和最广泛使用的方法。 校验和验证旨在通过综合分析 , 测试软件确保其正确运行功能,确保软件的质量和软件的可靠性。在 IEEE610.12(1990)中 , 测试被定义为在规定条件下对执行的系统或者组件进行观察和记录 , 并对系统或者组件进行评价的活动。

测试在实践过程中被广泛的使用, 在保证质量策略的诸多组织中扮演着重要的角 色。 软件影响着成千上万人的日常生活, 担负着艰巨的任务。 因此软件在不久的将来 将显得尤其的重要。研究表明,测试通常消耗软件开发成本的 30%至 50%。对于安全危急系统, 甚至更高的比例也不足为奇。 因此软件测试具有挑战的就是寻找更多的有效途径进行有效的测试。

软件测试管理的价值在于努力减少测试成本和满足需求。有价值的测试管理对于 项目目标和商业价值也能有很好的向导。 在第一章, Boehm列举了很多方面的潜在测试成本。该例子说明了利用客户结账类型的 7%的成本来完成 50%的软件测试利益。 尽管百分百测试是一个不太切实际的目标, 然而通过调整测试方法, 仍有很大的空间 来改进和节省达到预期的价值。 基于软件工程的价值动力在于目前软件工程的实践研究都是把需求, 测试案例, 测试对象和产品缺陷看的同等重要。 这对测试显然是正确的, 它对产品价值间接的做出贡献。开发和测试的分离使得这问题显得更为突出。 测试往往是一个纯粹的技术问题, 将使得测试和商业决策之间曾解链的关系变的更加紧密。 本章主要描述了提高基于价值的测试管理的需求 , 解释其基本要素, 讨论现有的实例来支持基于价值的测试 , 以及基于价值的测试管理的基本框架。 本章接下来部分 的基本结构如下:11.2节讨论测试管理的贡献价值; 11.3节讨论已测案例对于测试 管理的支持; 11.4节用例子描述基于价值的测试管理的框架。本章最后将对具体的再作更进一步的研究。

9文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

基于价值测试的描述

基于价值的校验和验证的目标被定义为确保软件的实施能够满足其预期的目标 价值利益。 如果我们从价值的角度去考虑, 那么测试的贡献应该在哪呢?从根本上来说 , 可以从两个方面来考虑: 内部方面包括测试成本以及测试效益 , 外部方面强调 未来系统的威胁以及机遇。对于基于价值的测试, 其关键就是要把这两个方面结合起来 , 也就是说通过客户和市场需求调整内部测试的过程。

为了使内部和外部两方面因素相结合,只专注于技术方面的测试显然是不恰当的。相反,测试管理需要有个全局的把握。

例 41描述了测试管理内外双方面之间的依赖关系。内部方面的就如同测试经理 对整个项目的控制。 这方面的费用主要来自于软件测试实践以及短期长期的测试。 而外部方面被认为是测试主管所能控制以外的一些利益和参数数据。基于价值的软件测 试管理使得测试能够满足以利益为重的价值主张, 以及使整个小组聚焦在有价值的测 试方向上。

对于软件测试的外部观点的首要问题是“我们如何确保软件系统的价值目标?” 这目标就是通过协调价值主张来 测试软件集中的有价值部分,最重要的品质以及项目风 险的及时调整等。回答这样的问题包括市场机遇,项目的价值主张以及成本效益。参考第一章关于机遇和风险的详细介绍以及参考第七章的价值主张的引出与调和。 内部观点是建立在价值主张的利益之上, 以及测试的预算代表着整个项目的一个大概水平。 这主要的问题就是如何把测试作为一项投资活动。 为了能够高效迅速的测 试及降低开发预算。适当的内外部交流协调能够满足测试的利益价值。

测试的价值贡献

10文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

测试与其他的开发环节诸如代码和用户界面设计相比, 它不能立即对产品创造价值。 然而测试提供和支持软件开发过程中产生的有价值的任务活动。 理解测试贡献价值的关键点在于测试的贡献效应。 测试的贡献建立了测试与最终产品价值利益之间的关系。

最直接的客户是直接与测试小组有密切联系的程序开发者和项目主管。在基于价值的软件工程测试过程中的集中力量是顾客和用户(见第七章) 。顾客和用户通过设 定语境和范围来进行测试达到测试的价值目标。

测试的客户

开发人员 , 项目主管 , 质量主管 , 顾客 , 分析者 , 最终用户或者维修人员们都得益于 软件系统的分析 , 依靠反馈来检测问题 , 降低其不确定性 , 做出相关的决定来加快产品 进程。

下面的例子显示了不同组对于测试需求的反馈信息:

顾客和用户关于多大程度上需求一致是否满意以及在多大程度上满足软件的价值利益。测试还对项目的进程提供可见性和洞察力。通过测试的结果可以了解已通过的测试案例 . 当验收测试时出现不实用的或者失败的显示在实际环境中才能出现的问题, α和 β测试提供了一个更加坚实的基础来验证结果。∙销售和产品主管从测试计划、 定价、 促销和分配方面获取相关的信息。 产品实际的质量与顾客和用户所期望的质量之间的差异很容易会导致误解和错误的 设想以至于降低或者阻止了真正价值的实现。为了能够成功地达到这些期望以及满足个人或组织的目标,通过客户需求来调整产品设计满足某些功能。 对于项目主管, 测试支持了风险管理和项目进程的估计。 重点是识别和排除潜 在的价值破坏和抑制价值功绩的风险。早期的大幅降低项目绩效的严重缺陷 , 是一个主要的目的。 测试

11文档收集于互联网，如有不妥请联系删除.

文档来源为:从网络收集整理.word版本可编辑.欢迎下载支持.

降低了不确定性和帮助项目主管对于清除缺陷、系统 稳定性以及产品更新发布能够做出更好、更明智的决定。

质量主管对于问题的识别以及对特定问题的动态趋势较为感兴趣。 测试结果对 于项目评估、对于质量策略的保证以及进程的改进提供了帮助。 Rosenberg 讨论了测试如何对确保质量做出贡献并展示了测试问题如何验证如何修正以此来提升项目进程。开发者和用户了解相关问题的当前状态, 并且提供相关数据 来衡量及预测软件的质量和可靠性。

开发人员通常需要获取反馈信息来验证测试实施是否完整, 是否符合标准, 是否满足质量要求。为了保证稳定性, 测试提供了相关缺陷的详细信息, 提示测试失败的原因。除此之外,测试对于项目缺陷的改进做出反馈。例如,通过相 关的修改以后需要测试其是否随着相关的改动使原先的功能有所改动或者出 现衰退的情况,这些都是需要注意的。

对于需求工程师来说,测试对于验证和确认需求是很有价值的。 Weinberg 曾指出 “最有效的方法之一就是通过测试案例就象测试一个完整的系统来形成测试需求”。黑盒测试通过其需求帮助能够保证它们的完整性,准确性,透明性 及简明性等。因此测试能够提升要求且向着测试驱动的方向发展。

简言之, 测试能够通过降低计划的不确定性和风险性来提升利益, 做出相关的决 定, 努力控制把不必要的消耗减少到最低程度 (内部原因 )。 尤为重要的是, 它有助于 实现预期的价值利益。 这些利益的得来也并不是免费的, 测试的代价通常也是有意义的(外部原因 )。测试可以被理解为购买信息,也可被认为降低成本风险,减少不确定性的投资活动。 在成本和利益的投资上需要对测试需求做出相关的决定。 因此接下来的两个问题是:什么是测试成本?什么是价值活动的测试利益?

12文档收集于互联网，如有不妥请联系删除.