New Edition of Software Metrics

Many software projects are navigating blindly: no idea how complex the code really is, how many test cases would be necessary or when the budget will be used up. Simple metrics can be used not only to precisely estimate testing efforts, but also to predict error rates months in advance - to an accuracy of ten percent. The trick lies not in complex mathematics, but in measuring quantity, complexity and quality in a targeted manner and drawing the right conclusions.

Podcast Episode: New Edition of Software Metrics

Together with Manfred, I have been working on the second edition of our book “Software Metrics” (formerly “Software in Numbers”) and now it’s here! Metrics are an exciting topic. Once you have selected suitable metrics from the categories of quality, quantity and complexity, they provide a deeper understanding of the software - and often bring things to light that you might not have wanted to know. We take a look at what we can use metrics for, how to select them and how to integrate them into everyday life.

“If I have 10 test cases, but 500,000 lines of code, then it becomes pretty clear that it might not fit together” - Manfred Baumgartner

Manfred Baumgartner has over 35 years of experience in software development, specializing in quality assurance and software testing. He studied computer science at the Vienna University of Technology and then worked as a developer and test manager before establishing QA consulting services at Nagarro GmbH in Vienna in 2001. Baumgartner is a member of the ASQF e.V. executive committee and has written specialist literature on topics such as system testing and agile test techniques.

Highlights der Episode

• Software metrics are like a speedometer and fuel gauge - without them, you’re driving blind through the project.
• Trend more important than absolute value: old systems don’t have to be perfect, just not worse.
• Complexity determines effort more than quantity - ten complex test cases beat a hundred simple ones.
• Simple metric: evaluate business case by complexity, derive number of test cases, adjust by sprint - works.
• Agile projects neglect metrics because errors disappear into the backlog instead of being documented.

Deciphering software metrics

Software metrics are essential for understanding and improving software projects. Today I talk to Manfred Baumgartner about the importance of software metrics, their applications and how best to get started with them.

Why software metrics?

Today I spoke with Manfred Baumgartner. Together we have published the second edition of the reference book “Software Metrics” (formerly “Software in Numbers”). The relevance of software metrics cannot be emphasized enough - they are the backbone for understanding and optimizing the processes within a software project. Manfred talks about how metrics can be used in projects to not only measure performance, but also to make development and testing processes more effective.

The diversity of metrics

Software metrics are diverse and can be divided into different categories. Manfred emphasizes the importance of this diversity by referring to quantity, quality and complexity metrics. Each category highlights different aspects of a project and provides valuable insights into areas such as code volume, error density or even the architectural nature of an application.

Use cases of metrics

Manfred shares specific examples of how metrics have been applied in real projects. From measuring code quality to evaluating test coverage, each metric provides crucial data points for project managers and development teams. The role of metrics in decision making, such as selecting technologies or determining the development effort for new features, is particularly emphasized.

Challenges and solutions

Despite their obvious advantages, software metrics also encounter challenges. One of these is acceptance within teams - there is often a fear that metrics could reveal negative aspects. However, Manfred emphasizes that the goal is not to point out mistakes, but to identify opportunities for improvement. He also provides practical tips for getting started with software metrics and emphasizes the importance of targeted implementation.

A pragmatic approach

To fully utilize the benefits of software metrics, Manfred recommends a pragmatic approach. Instead of jumping straight into complex measurement procedures, teams should start with basic metrics and observe their effects. Through iterative adjustments, customized metrics systems can be developed that are specifically tailored to the needs of a project.

The future of software metrics

Manfred talks about the future of software metrics and their role in an ever-changing technological landscape. He foresees a growing awareness of the value of these tools and an increasing integration into the development cycle - a positive step towards more efficient and higher quality software projects.