<h2>Must know</h2>
<ul>
<li><strong>Google mandates post-quantum migration by 2029</strong>: Citing rapid advances in quantum hardware and the ‘Harvest Now, Decrypt Later’ threat, Google has set a 2029 deadline for full migration to Post-Quantum Cryptography (PQC). Android 17 will lead this shift by embedding ML-DSA digital signature protection. Security experts via industry reports urge organizations to conduct immediate cryptographic inventories and adopt crypto-agile infrastructure to swap algorithms as NIST standards evolve. <a href="https://app.daily.dev/posts/2Ya2QAFm3" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>TypeScript 7.0 pivots to Go-based compiler</strong>: The release of TypeScript 6.0 marks the final version built on its original JavaScript codebase, with version 7.0 set to introduce a new compiler written in Go. This transition, codenamed Project Corsa, aims for native speed and multi-threaded type checking, with early benchmarks showing a 10x reduction in check times for massive codebases. Developers are advised to upgrade to 6.0 now to handle deprecated legacy options like AMD/UMD targets before the architectural shift. <a href="https://app.daily.dev/feed-by-ids?id=2TBDrCEMI&amp;id=fwC1fC7w1" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AppArmor exploit enables root execution via su and sudo</strong>: Qualys researchers discovered a sophisticated Linux privilege escalation vulnerability (CVE-2026-001) that chains the ‘confused deputy’ problem across setUID binaries. By using <code>su</code> to write to AppArmor’s kernel pseudo-files, an unprivileged user can prevent <code>sudo</code> from dropping root privileges, eventually gaining full root code execution. Immediate updates for AppArmor and <code>sudo</code> are required to mitigate this elegant but dangerous exploit chain. <a href="https://app.daily.dev/posts/a8sN8k1SL" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>AWS Lambda expands to 32GB RAM and 16 vCPUs</strong>: New configurations allow custom memory-to-vCPU ratios for compute-heavy workloads like media transcoding. <a href="https://app.daily.dev/posts/jP3XPTAuv" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Betterleaks replaces entropy with BPE token scanning</strong>: The creator of Gitleaks launched a new Go-based secrets scanner designed to catch ‘vibe coding’ credential leaks. <a href="https://app.daily.dev/posts/CnjLBDMza" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Wine 11 delivers massive Linux gaming gains</strong>: A kernel-level NTSYNC thread synchronization rewrite significantly improves performance for Windows games on Linux. <a href="https://app.daily.dev/posts/pNy5xKg0N" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cloudflare Workflows adds AST-based visual debugging</strong>: The system uses a Rust-based parser in WASM to statically derive execution graphs from TypeScript code. <a href="https://app.daily.dev/posts/2kRwcIErj" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Domscribe gives AI coding agents eyes on the DOM</strong>: This open-source tool uses AST transforms to map live frontend elements directly to source file lines for AI agents. <a href="https://app.daily.dev/posts/BCgkhGZVL" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>LLVM DTLTO link times drop by 90 percent</strong>: A new patch in LLVM 23 enables efficient file addition via memory buffer ownership, bypassing slow disk reads. <a href="https://app.daily.dev/posts/D8gm3FYAM" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Velxio emulates 19 embedded boards in the browser</strong>: A new AGPLv3 project supports AVR8, ARM, and RISC-V architectures with interactive electronic component simulation. <a href="https://app.daily.dev/posts/mMt2iwRIy" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>The vibe coding spec-driven development paradox</strong>: While some ‘thought leaders’ push ‘vibe coding’ as the future, others like Gable’s Head of DevRel are now rebranding basic engineering—writing specs and diagrams—as ‘Spec-Driven Development’ for AI. It turns out that to make AI not ship garbage, you have to do all the senior engineering work you were trying to avoid in the first place. We’ve successfully automated the coding part just to spend twice as much time writing JSON schemas for the bot. <a href="https://app.daily.dev/feed-by-ids?id=LFFkW7oNN&amp;id=CnjLBDMza" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Cursor’s frontier model is just someone else’s open source</strong>: Cursor launched ‘Composer 2’ with the ‘frontier model’ label, conveniently forgetting to mention it’s actually built on Moonshot AI’s open-source Kimi K25. It’s a classic Silicon Valley move: wrap an open-source model in a slick UI, slap a vague marketing term on it, and hope nobody asks about the licensing or attribution. Transparency is apparently the first thing to go when you’re chasing the next funding round. <a href="https://app.daily.dev/posts/460WsqsI4" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Palantir’s Maven kills humans while the media blames LLMs</strong>: After a US strike hit an Iranian school, headlines rushed to blame Claude, but the actual culprit was Palantir’s Maven Smart System and its 1,000-decisions-per-hour ‘kill chain.’ It’s much easier for the press to freak out about ‘charismatic’ LLMs than to address the systemic failure of high-speed targeting databases that remove human deliberation. We’re blaming the chatbot for the sins of the military-industrial complex’s CRUD apps. <a href="https://app.daily.dev/posts/jaZlJ2Voo" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
