<h2>Must know</h2>
<ul>
<li><strong>Vite 8 and Rolldown signal the end of the Babel era</strong>: Vite 8 has officially launched, replacing Rollup and esbuild with the Rust-based Rolldown for 10-30x faster builds. This release includes a Babel-free React plugin and Wasm SSR support, significantly reducing the toolchain complexity for frontend engineers. JavaScript Weekly reports that this move, alongside the open-sourcing of the Vite+ unified toolchain, aims to consolidate linting, formatting, and bundling into a single high-performance workflow. <a href="https://app.daily.dev/feed-by-ids?id=db4oA9qt6&amp;id=V8fBkmgAD&amp;id=J1Rw9aL2Z" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Mistral Forge challenges OpenAI with full model training</strong>: Mistral announced Forge at Nvidia GTC, a platform that allows enterprises to train custom models from scratch rather than relying on RAG or fine-tuning. This is a strategic pivot toward ‘agent-first’ architectures where autonomous systems manage their own training pipelines using proprietary data. Mistral claims this approach, supported by partners like ASML and the European Space Agency, provides better handling of domain-specific codebases and non-English data. <a href="https://app.daily.dev/posts/6pejPwPmX" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GNOME 50 Tokyo drops X11 and embraces Wayland only</strong>: The GNOME 50 release marks a definitive shift by entirely removing X11 support in favor of a Wayland-only environment. Key improvements for developers include official fractional scaling, VRR support, and a new Vulkan-accelerated remote desktop backend. According to GNOME Release Notes, the update also brings significant NVIDIA performance fixes and will be the default for the upcoming Ubuntu 26.04 LTS. <a href="https://app.daily.dev/posts/Xg0Exf55t" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Edge.js runs Node apps in a WebAssembly sandbox</strong>: Wasmer’s new runtime offers full Node v24 compatibility inside a WASIX sandbox with only a 5-30% performance hit. <a href="https://app.daily.dev/posts/FUA9hYuPV" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Bun v1.3.11 adds native cron and ANSI string slicing</strong>: The update introduces Bun.cron for OS-level scheduling and fixes over 100 bugs, including critical Windows ARM64 shims. <a href="https://app.daily.dev/posts/DuG54Q8tB" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Zeroboot achieves sub-millisecond VM forking for AI agents</strong>: Using Firecracker and copy-on-write, this tool forks sandboxed microVMs in 0.8ms, crushing existing benchmarks. <a href="https://app.daily.dev/posts/0KZTYS81s" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Static analysis of 500 repos reveals massive memory leaks</strong>: A study found 86% of React and Vue repos leak memory due to missing cleanup in useEffect and onMounted hooks. <a href="https://app.daily.dev/posts/b6K3Wc80v" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Arcjet launches WebAssembly-based prompt injection protection</strong>: The new SDK blocks structural exploits like ChatML injection at the application lifecycle level for Next.js and Python. <a href="https://app.daily.dev/posts/B5yvripTT" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>ConnectRPC compiles GraphQL into typed SDKs</strong>: Cosmo Router now supports generating gRPC and typed TypeScript clients directly from named GraphQL operations. <a href="https://app.daily.dev/posts/Cgx6Z8AyK" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Fedora Asahi Remix 43 brings 120Hz to Apple Silicon</strong>: The latest release supports M1/M2 Mac Pro models and enables high-refresh rates for MacBook Pro 14 and 16. <a href="https://app.daily.dev/posts/585WUMr9y" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>The 19% productivity tax of AI coding</strong>: While marketing departments scream about 10x developers, a study by SmartBear and independent researchers found that AI tools actually cause a 19% overall slowdown for many. Developers are spending 50% more time debugging ‘plausible-looking’ garbage code, effectively turning senior engineers into high-priced janitors for LLM hallucinations. It turns out that generating boilerplate at the speed of light doesn’t help when you spend the afternoon chasing a cache isolation bug the AI missed. <a href="https://app.daily.dev/feed-by-ids?id=Lb19kFYlj&amp;id=cY9Cx1sh3&amp;id=ab5QxLxXV" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft superintelligence is the new pivot to AI</strong>: Mustafa Suleyman is ‘stepping back’ from the struggling Copilot product to focus on a ‘superintelligence initiative’ that will supposedly bear fruit in five years. This is classic corporate-speak for moving a high-profile hire away from a product that isn’t hitting market targets despite being forced into every Windows context menu. If you can’t make a chatbot summarize an email reliably today, just promise a god-in-a-box by 2031 and hope everyone forgets the current PR backlog. <a href="https://app.daily.dev/feed-by-ids?id=jTMZOA2Fr&amp;id=w3Je7Auua" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>The Arena leaderboard is a circular firing squad</strong>: Arena (formerly LM Arena) has reached a $1.7 billion valuation as the ‘unbiased’ gold standard for LLM rankings, despite being funded by OpenAI, Google, and Anthropic. We are expected to believe in the ‘structural neutrality’ of a leaderboard where the judges are paid by the contestants. It’s a perfect hype loop: pay for the benchmark, optimize for the benchmark, and then use the benchmark to justify your next $10 billion compute raise. <a href="https://app.daily.dev/posts/exy1uRyL2" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
