Sponsored by Broadcom
To improve the developer experience, it’s important to begin by understanding those with the most important perspective on it: the developers.
In recent years, developer experience (DevX) has been the subject of extensive research by academics and small and large private organizations such as Microsoft and Stack Overflow. This trend began a decade ago with the emergence of Agile and DevOps practices coupled with the desire to drive higher levels of developer productivity and satisfaction to meet the needs of modern digital businesses.
This research has sought to answer some of the most challenging and fundamental questions on attitudes, preferences around tools and practices.
- How do we reduce friction and optimize for automation?
- How should developers incorporate new shift-left responsibilities?
- How do we keep developers in the “flow state” — those windows of maximum productivity?
- How does organizational culture impact developer happiness and motivation?
- And the newest addition: How can we use generative artificial intelligence (GenAI) to improve productivity and overall delivery velocity?
Surveying Mainframe Developers
While many research studies have canvassed developers from diverse backgrounds, resulting in invaluable insights and recommendations, the voice of mainframe developers, a small but important segment of the global community, has been largely drowned out. Their experiences and challenges are unique in many ways, so we set out to understand their perspective — the good and the bad — specifically.
Broadcom commissioned a study by Forrester Consulting of over 800 IBM z/OS application developers across organizations, geographies, and generations. The study asked questions to discover their likes and dislikes, their biggest challenges, and how they would speed software delivery. There’s an added sense of urgency due to GenAI’s growing disruption to existing standards and practices. How would developers adopt this technology to improve the DevX and drive new levels of application vitality?
Let’s look at the study’s findings and wide-ranging recommendations.
How Is Mainframe AppDev Different?
One of the challenges voiced by developers stems from the mainframe’s tendency to run extremely large, complex applications that have been maintained over extended periods of time. As a result, code comprehension is difficult. Over half of the respondents reported needing three days or more just to understand where to make code changes.
The good news is that there are opportunities, both current and emerging, to tackle the comprehension challenge. For example, with modern IDEs (Integrated Development Environment) like VS Code (Visual Studio Code), developers can run regular expression searches across entire folders which greatly facilitates code navigation (especially when the alternative is endless page-downs with ISPF). Also, the new cloud IDEs open the door to more collaborative coding (e.g., pairing knowledgeable veterans with new-to-mainframe talent to facilitate knowledge transfer) and the GenAI “code explain” function is rapidly advancing for languages like COBOL.
Changing Demographics
While mainframe AppDev has a history of being isolated from other teams, 56% of respondents now have professional experience developing on platforms other than mainframe. When you combine this statistic with those whose education was based on other platforms and those who practice as citizen coders or hobbyists using other platforms, the number is likely much higher, reflecting a clear tipping point: more mainframe developers are familiar with contemporary tools and practices than not.
Biggest Barriers to Productivity
Unsurprisingly based on the previous finding, the top reported barrier to productivity is “outdated tools and frameworks.” Given the large number of developers with experience on other platforms, it’s not a surprise that the use of contemporary tooling like VS Code, Git, and automated DevOps toolchains would become table stakes. In fact, 70% of respondents believe VS Code adoption would result in a major increase in productivity while 74% say the same about Git.
“Interruptions/context switching” was the second biggest productivity barrier. It’s clear these developers don’t perceive their day-to-day experiences as streamlined or friction-free — there’s plenty of room for improvement.
The third biggest barrier was a tie between a “lack of autonomy” and a “lack of Agile/DevOps practices.” The former can be addressed with more self-service options for testing, databases/environment set-up, etc. The “lack of Agile/DevOps practices” barrier may simply require coordination with enterprise IT and DevOps champions to understand and evaluate internal corporate standards. Those familiar with Agile/DevOps practices and tooling on the mainframe team can facilitate adoption, acting as internal advocates and trainers.
Time Spent Coding
This is a good news, bad news story. The surveyed developers spend more time coding than any other activity (the good news), but it’s only 16% of their time (bad news considering 30% is a commonly cited benchmark). The interruptions and dependencies mentioned above illustrate just how disjointed mainframe development is today. Autonomy and time spent coding are generally positively correlated (i.e., more time coding, less time coordinating with others).
Base: 838 global mainframe application developers at enterprise companies
Source: A commissioned study by Forrester Consulting on behalf of Broadcom, June 2024
High Level of Happiness
The survey reflected a satisfaction rate of 93%, with 68% saying they were extremely satisfied with their careers. Recommending their career is another measure of happiness, and 86% of respondents are “very likely” to recommend it to others.
When asked about the reasons for their satisfaction, respondents cited:
- High-performance platform
- Professional opportunities
- Job stability, and,
- Access to cutting-edge technology
How to Accelerate Software Delivery
The most common recommendations were to focus on automation while staying connected to the rapid advancements in GenAI enablement which, when you consider the biggest productivity barriers, makes sense. These developers understand the automation opportunity and are hungry to eliminate manual, repetitive tasks.
In summary, by engaging and listening to developers, we can forge a new, highly automated experience that unlocks their creativity while driving a new level of code change confidence and application vitality.
To download the complete Forrester study, visit the Developer Survey site.