<h2>Must know</h2>
<ul>
<li><strong>Snowflake embraces open data standards with pg_lake integration</strong>: Snowflake has integrated <code>pg_lake</code>, an open-source PostgreSQL extension, into its Snowflake Postgres service. This allows users to work with data lakehouses and Apache Iceberg tables directly via SQL, with automatic versioning, data type conversions, and caching. This move, along with expanding its Horizon Catalog for Iceberg tables, emphasizes interoperability and open standards, aiming to unify analytical and transactional workloads and prevent vendor lock-in, according to Snowflake. <a href="https://app.daily.dev/posts/kBWxo6cQz" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Microsoft unveils rust-based sandboxing project litebox</strong>: Microsoft open-sourced LiteBox, an experimental Rust-based library OS for application sandboxing. It offers stronger isolation than containers without the overhead of full VMs, using a split architecture for POSIX system calls and host connections. While not production-ready, it aims to enhance security for applications needing robust isolation and supports confidential computing with AMD SEV-SNP. <a href="https://app.daily.dev/posts/jIlb9nmzq" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Apple’s lockdown mode thwarts FBI data extraction</strong>: The FBI’s Computer Analysis Response Team (CART) was unable to extract data from a Washington Post reporter’s iPhone due to Apple’s Lockdown Mode being enabled. This incident, reported by Schneier on Security, occurred during a raid related to a classified information leak investigation, highlighting the effectiveness of iPhone security features against law enforcement forensic tools and raising implications for press freedom. <a href="https://app.daily.dev/posts/PXLqFkE9O" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>MUI launches base UI 1.0 with 35 accessible components</strong>: MUI has released Base UI 1.0, an unstyled React component library offering 35 accessible components with full styling control and WCAG compliance, backed by a dedicated team for long-term maintenance. <a href="https://app.daily.dev/posts/lFpG64iAY" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>State of javascript 2025 survey results released</strong>: The State of JavaScript 2025 survey results are out, providing insights into the latest trends, tools, and community preferences in the JavaScript ecosystem. <a href="https://app.daily.dev/posts/lJAPeqJsk" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Interop 2025 achieves 97% pass rate across browsers</strong>: Interop 2025 concluded with a 97% pass rate across all browsers, showing significant progress in web platform interoperability, with Safari making the largest improvement. <a href="https://app.daily.dev/posts/X7hlaiKma" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Ubuntu expands RISC-V support with K3 and K1 processors</strong>: Ubuntu is expanding its RISC-V support, targeting the new SpacemiT K3 processor with Ubuntu 26.04 LTS and bringing Ubuntu 24.04 LTS to the older K1 chip, strategically important for IoT and edge computing. <a href="https://app.daily.dev/posts/C9IUbBaRM" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Qodo 2.0 enhances AI code review with specialized agents</strong>: Qodo launched version 2.0 of its AI-powered code review platform, introducing multiple specialized AI agents that use recall and memory to review code at a senior engineer level, aiming to maintain code quality amidst increasing AI-generated code. <a href="https://app.daily.dev/posts/PAc5sU2Jg" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>GitHub improves pull request file change experience</strong>: GitHub has updated its pull request “Files changed” experience, adding CODEOWNERS validation and significant performance improvements, making navigation faster and reducing memory usage for large pull requests. <a href="https://app.daily.dev/posts/XX00GiT1G" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>PureRef 2.1 brings new features and language support</strong>: PureRef released version 2.1, its first update since 2024, introducing a background grid with snap-to-grid, shape drawing tools, and support for six new languages. <a href="https://app.daily.dev/posts/EOJeYNj9E" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI-accelerated AWS breach: same old story, faster</strong>: Apparently, AI is so good now that it can breach an AWS environment in 8 minutes, escalating to full admin privileges and exfiltrating data, all thanks to LLMs. The real culprit? Misconfigured S3 buckets. So, AI is just really good at finding the same old human screw-ups, not some magical new attack vector, according to AWS. <a href="https://app.daily.dev/posts/6mJxFRcx6" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Operant AI launches agent protector for enterprise AI security</strong>: Operant AI has launched Agent Protector, a real-time security platform to monitor and control AI agents in enterprise environments. Because, you know, autonomous AI agents accessing sensitive data and operating rapidly is a <em>new</em> problem that definitely requires a whole new product, not just, you know, basic security hygiene and access controls. But hey, 40% of enterprise apps will have AI agents by late 2026, so better buy something! <a href="https://app.daily.dev/posts/RMh1riegC" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI model race heats up with new releases and dubious achievements</strong>: Anthropic released Claude Opus 4.6 and OpenAI released GPT-5.3 Codex within an hour of each other, sparking a “fierce competition” for enterprise AI. One developer even used 16 Claude Opus 4.6 instances to build a 100,000-line Rust C compiler in two weeks for a mere $20,000. Because nothing says innovation like throwing a bunch of AI at a problem with “ideal conditions” and then calling it a breakthrough, completely ignoring the human skill still required to define the problem and verify the output. <a href="https://app.daily.dev/feed-by-ids?id=6vLfIe4vv&amp;id=8HMA0hoKL&amp;id=B4HtIccoU&amp;id=BQ8poB6iB&amp;id=CWUtlxYzI&amp;id=D8NcAVvwq&amp;id=EmRFinQYK&amp;id=LVl8sC9Gk&amp;id=YIoiLTXiE&amp;id=gP03wq1GE" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
