Kevlin Henney

🐌 Is QA slowing you down? Or is the real issue a lack of shared visibility and structure? In this article, Tiago Moura Evangelista shares how years of dealing with fragmented processes led to one clear conclusion: when QA operates in isolation, it becomes a bottleneck. By rethinking the role of QA and adopting a structured approach with SquashTM, his team was able to: Align requirements, tests, and results in a single flow Provide clarity to all stakeholders, early and continuously Shift QA from a gatekeeper to a facilitator of confident releases 👉 Read the full article: https://lnkd.in/ey_SyR_t Because when quality information flows, teams move forward ! #QA #SoftwareTesting #TestManagement #SquashTM #QualityEngineering #ProductDelivery #ReleaseConfidence

5

Will AI coding agents replace software developers? It’s the hottest debate in tech right now. Some say developers will disappear, others say nothing will change. My view sits somewhere in the operational middle. Before forming an opinion, it’s worth observing something interesting. 🤖 The People Closest to the Models Are Often the Most Realistic A pattern I keep noticing: • The strongest believers in “AI replacing developers” are often far from real software development • The engineers working closest to the models tend to be far more measured • Many see the limitations very clearly This isn’t skepticism, it’s proximity to reality. 💡 Writing Code Is Design Thinking Coding is often misunderstood as typing lines of syntax, in reality it’s design. When developers work, they are: • Defining system boundaries • Managing trade-offs • Anticipating failure modes • Designing interactions between components That’s architecture thinking. LLMs, by contrast, are probabilistic tools, they predict the next token based on patterns in training data. That’s powerful, but it’s not the same as strategic reasoning about a system. They can remix what exists, yhey don’t truly understand the system they’re designing. ⚠️ Reliability Still Matters Another practical reality - LLMs will ALWAYS hallucinate. That’s manageable in many contexts, but production systems have a very low tolerance for mistakes. When code runs critical pipelines, financial systems, or operational processes: • A small error can cascade • Edge cases matter • Assumptions must be explicit That responsibility still sits with engineers. 🚀 So What Actually Happens? AI will absolutely change software development. Developers using AI tools will be dramatically more productive. • But the role doesn’t disappear, it evolves. • Less time writing boilerplate, more time designing systems, more time validating outcomes. In other words: The value shifts from typing code to thinking about systems. And that was always the hall mark of a good engineer anyway!

35

25 Comments

Are your Go dependencies becoming a tangled mess? 🧹 Time to spotlight another excellent talk from GopherCon UK: "Go Module Hygiene: Keeping go.sum and go.mod in check." ✅ 🔎 In this practical session, DevOps and Software Engineer Emily A. dives into the unsung heroes of your Go projects: the `go.mod` and `go.sum` files. She explains how, without proper care, they can quickly bloat with unused packages, outdated versions, and potential security risks. 🛡️ The talk provides actionable strategies to keep these files clean and efficient. Emily covers essential techniques for identifying and removing unnecessary dependencies, resolving version conflicts, and auditing modules to safeguard your projects and maintain reproducible builds across all environments. 🎙️ The full recording is now live on our YouTube channel: https://lnkd.in/dt94zX8b #GopherConUK #Golang #Go

57

And now for some readings of user level RCU .. and a landmark paper that led to the implementation of liburcu - the user space RCU library, a Christmas Day evening read .. Why RCU is difficult in user space ? RCU, particularly its high-performance Quiescent-State-Based Reclamation (QSBR) variant, is easier in kernel mode because the kernel scheduler automatically detects quiescent states whenever a CPU context-switches, enters user mode, or idles, allowing grace-period tracking without any explicit cooperation from kernel code. In user mode, applications lack this built-in mechanism, so threads must explicitly register and periodically report quiescent states (e.g., by calling specific functions), imposing invasive global constraints on the entire application. These requirements make user-level RCU harder to adopt broadly, as they complicate library design and require modifications to all potentially reading threads, which is impractical in many user-space programs. User level implementations of RCU : This paper contributes to user-level RCU by formally describing efficient and flexible implementations that overcome the limitations of prior approaches, which either imposed high read-side overhead or severely restricted application design. It presents multiple classes of RCU (including QSBR, memory-barrier, signal-based, and bullet-proof variants) with detailed algorithms, performance analysis, and comparisons to locking, directly forming the foundational basis for the liburcu library's core flavors and enabling its widespread adoption in user-space applications.

74

3 Comments