Fork your dependencies, trim them to only your use case, never update unless it breaks for your users. I’ve been vocal about this for 10 years. I’ve always said that updating is way riskier than latent bugs (which can be tracked and CVEs monitored). If you are updating a dependency, it’s on you to analyze every single commit in the full transitive set of dependencies. If you dont see anything compelling, dont update! I remember at HashiCorp once in awhile an engineer would try to update a dep or replace a DIY lib with an external one and id always ask “show me the commit we need.” Dont update for the sake of it. Feeling pretty swell about this mentality with all the supply chain attacks happening.

Blowlang apparently allows arbitrary code execution at compile time. While technologically impressive, it might be possible to get hacked, if you compile malicious code. Can anyone with access to the compiler verify this? Or disprove it?

I believe one reason for the bloat in modern applications is that we build them on powerful computers, which hide coding inefficiencies. Unfortunately, IDEs and development tools also demand top-tier machines because they consume significant resources. Although the comparison may not be entirely fair, I often think about the simple chat function of ICQ from two decades ago versus Microsoft Teams today. Teams offers far more features and built-in security, but ICQ still outshines it in basic chat performance and responsiveness. I remember running ICQ on an Intel 90 MHz single-core machine with 64 MB of RAM on Windows 95. It launched instantly and the chat functionality worked seamlessly. In contrast, Teams can take several seconds, sometimes even minutes, to start up and often freezes on my 64 GB, Intel 3.0 GHz, 16-core machine. I suspect many people feel the same frustration with modern software. So where did things go off track? ICQ was developed in 1996 under very tight constraints, which forced the developers to build an extremely efficient program. Any memory leak would have been immediately noticeable., the program simply wouldn’t run properly, forcing a fix. High CPU usage would cause the system to freeze, pushing developers to optimize or find creative ways to work within those limits. I often wonder how efficient modern applications could be if they were developed under similar constraints. Such limitations would push us to prioritize coding efficiency. Memory leaks or excessive CPU usage, issues that might otherwise go unnoticed on powerful machines,,would surface during development. I’m not opposed to high-end machines. But I do think developers could benefit from occasionally working in constrained environments. While increasing resource requirements may sometimes be necessary to unlock software’s full potential, I’m not convinced we truly understand where that threshold lies. Not to mention, it becomes difficult to stay productive on a standard computer with modern IDEs.

ngl, these headlines always mess with my head

Garbage collection is still unfit for demanding games: 1. GC requires ~2x the memory footprint to achieve equivalent performance to an game using manual memory management. 2. GC consumes extra threads and memory bandwidth that could've been used for the game. 3. ZGC trades off throughput for latency. In layman's terms, your game will run slower, but at least you won't have lag spikes.

NEW: A Cursor AI coding agent deleted a startup's entire production database in 9 seconds. The agent, powered by Claude, was working on a staging task, found a broadly scoped API token, and executed a volume delete without confirmation. It later confessed in detail, admitting it guessed and violated safety rules. PocketOS, which powers car rental businesses, lost months of bookings data. In short, the AI agent rouge.

Big Computer Science wants you to believe quick sort is the best sorting function. Turns out, they're right, but it's good to check anyway 🙂 Insertion sort peeks ahead at low values of N and when the array is nearly sorted, but not sure it's compelling enough to be the default option to reach for, imo.

Would you pay for a subscription to a "Windows Professional" if it included NO telemetry, ads, sponsored links, or upsells, bloat, or junk? Clearly, they have to make money. Today, however, it seems everything is predicated on monetizing the customer post-purchase. And that's largely because people don't pay for Windows. Put another way, if Windows were clean and didn't include any bloat or upsell, would you actually pay for it?