[$] Last rites for Gentoo's Chromium package
Gentoo is ending support for its Chromium package due to the unsustainable "packaging tax" imposed by its complex build system and frequent releases. The maintainers cited the burden of managing bundled dependencies and constant user complaints as the primary reasons for the project's "last rites."
LWN · 09 October 2026 · Read the original →
[$] An update on Rust's project goals
The Rust Foundation is implementing a formal goal-setting process to align institutional resources with technical priorities like "Rust for Linux." This initiative aims to ensure that high-impact projects have the necessary organizational support and contributor alignment to succeed.
LWN · 09 October 2026 · Read the original →
[$] An update on the Sashiko patch-review system
The Sashiko system has transitioned from an experiment to a functional part of the Linux kernel's development, using Large Language Models to automate initial patch reviews. This effort aims to address the chronic "review bottleneck" that limits the project's ability to process incoming code.
LWN · 09 October 2026 · Read the original →
Open Enables a Lot: The OSI at UNGA81
The Open Source Initiative is engaging with the UN General Assembly to define the role of Open Source AI in digital sovereignty and public infrastructure. These discussions are critical for establishing how international regulations will treat open models and the infrastructure they support.
Open Source Initiative · 09 October 2026 · Read the original →
The Acceptance Gap: Why AI-Generated Code Still Fails to Become Shipped Work
A survey of software developers indicates that 42% of committed code is now AI-generated or assisted, with that figure expected to rise to 65% by 2027. This shift is creating an "acceptance gap" where the volume of generated code outpaces the ability of teams to verify and ship it.
Tidelift · 09 October 2026 · Read the original →
What Skyscanner's AI rollout proves about verifying AI-generated code
Skyscanner's rollout of AI coding agents reveals that the primary challenge is not code generation but the erosion of trust when engineers stop reading every line of code. The organization is moving toward automated verification layers to catch mistakes that human reviewers no longer see.
Tidelift · 09 October 2026 · Read the original →
Turns out writing the code was never the hard part.
6 stories, every Friday
Published here every week. Follow by RSS to get it as it lands.