Interview with the New ACM TOSEM EiC, Prof. Abhik Roychoudhury

Interview conducted by Aldeida Aleti and Cristian Cadar
ACM TOSEM Information Directors

Prof. Abhik Roychoudhury is the new Editor-in-Chief for ACM Transactions on Software Engineering and Methodology (TOSEM) since January 2025. In this short interview, we ask him about his thoughts on the software engineering field and the latest initiatives for TOSEM.

The Software Engineering discipline is (at least) 50 years old. What lessons from past decades in software engineering are most relevant today?

TOSEM is the premier journal on software engineering and methodology. Software engineering is a core computing discipline providing a powerful mix of science and engineering. Software engineering is often mistakenly associated with programming but it is so much more than that, with important lessons on requirements, maintenance, evolution, analytics and human aspects. The lessons from these fields remain crucial, as we embrace the new wave of automation in software engineering. I have highlighted these aspects in the TOSEM editorial published in January 2025, as the incoming Editor-in-Chief. For example, as the wave of automation sweeps over coding, lessons in requirements elicitation and requirements engineering remain crucial. Similarly, as automatic programming becomes commonplace, the discovery of developer intent — not just for the whole software system, but also for its modules / functions — remains important. Thus, I feel the foundational concepts in software engineering will be adapted, and yet will continue to inspire us.

What are the new initiatives for TOSEM in the coming years?

TOSEM is a flagship journal of software engineering, and it is thus extremely important for TOSEM to stay in touch with the entire software engineering community. To further enhance the community outreach and promote strong discussions with the software engineering community — we will be holding an annual workshop. There was a 2030 Software Engineering workshop on the software engineering roadmap that was held in 2024. Instead of a one time event to craft a roadmap for software engineering, we hope to continue it as an annual event to engage and interact with the community. Furthermore, as software engineering faces new waves of disruption owing to automation, it is even more important to promote discussions on the evolution of our discipline. This annual workshop will take place every year with the ACM SIGSOFT Conference on Foundations of Software Engineering (FSE).

Furthermore, at the annual workshop, we plan to start giving out awards for the best papers in TOSEM published five years ago. The awards will be given in FSE 2025 in Norway for the first time to selected journal-first papers published in 2019–20. We choose the five year timeline after due thought and consideration with the TOSEM editorial board. It is to recognize medium-term impact, and draw the attention of the community to papers which are showing traction, so that we can have more discussions and reflections on those papers.

Last but certainly not the least, we have set up an Information Directorate headed by Aldeida Aleti and Cristian Cadar — to have the TOSEM editorial board interact with the broader software engineering community. Even this blog post is a result of the efforts from this Information Directorate !

The number of submissions to TOSEM have grown substantially over the past few years. What are the plans for scaling the TOSEM reviewing process without compromising quality?

Indeed, TOSEM has grown tremendously in the last five years in terms of submissions. To cater to the ever growing number of submissions, while maintaining review quality — we clearly face a scale-up issue. A key issue is the alignment of reviewers for a paper to the core topic or contribution of the paper, as imagined from the author’s perspective. For this reason, for regular papers submitted to TOSEM we have now changed the submission form so that authors can choose a primary area from eight areas:

• AI & Software Engineering
• Security & Software Engineering
• Requirements & Architecture
• Maintenance & Evolution
• Analytics & Empirical Studies
• Human & Social Aspects
• Testing, Analysis & Verification
• Others

Based on the choice of the primary area by the author, the paper is handled by a Senior Associate Editor, who along with an Editor oversees the review process. This to align the review process a bit more closely with the authors’ planned contributions!

How can the software engineering community balance foundational research with rapidly emerging technologies? In particular, how can generative AI improve software engineering?

This is really open to discussion and interpretation and our community members have different viewpoints, as indeed we should. In the end, a lot also depends on what we mean by an AI Software Engineer and what capabilities it can bring to the table? Is it only micro-capabilities like fixing a bug, or is it also capable of macro-capabilities like taking over somebody else’s code, understanding it, and writing new tests? Moreover, I notice that a lot of the discussions today are on an individual software engineer’s task automation. But software engineering is typically not a lone-wolf activity, team dynamics are involved in software engineering. So how do the diversity of thought and team interactions figure when we have AI as a member of a software team in the future ?

For these and other reasons, I continue to feel that the software engineering principles capturing requirements, design, testing, analysis, communication — will continue to play a pivotal role in the software engineering of the future, even if we see significant automation in software engineering tasks.

This is not to say that there are no broad changes. One of the changes that may be happening is the shifting of balance from scale to trust. Ever since the first Hello World program was written by Kerninghan in 1972, we have seen software engineering in-the-large with millions of lines of code being common in software projects. A lot of the focus has been on the management of scale. However with programming becoming automated, automatically generated components may increasingly get integrated into code-bases. In that case, the focus may shift from programming at scale, to programming with trust — where we focus on trustworthy integration of automatically generated software!

Disclaimer: The posts in the SIGSOFT Blog are written by individual contributors and any views or opinions represented in their posts are personal, belong solely to the blog authors and do not necessarily represent those of ACM SIGSOFT or ACM.