<h2>Must know</h2>
<ul>
<li><strong>CISA warns of active exploitation in VMware vCenter flaw</strong>: The Cybersecurity and Infrastructure Security Agency (CISA) has confirmed active exploitation of VMware vCenter Server vulnerability CVE-2024-37079, a critical heap overflow flaw with a CVSS score of 9.8. This vulnerability allows remote code execution and can be combined with other exploits to gain root access to ESXi. Broadcom released a patch in June 2024, and federal agencies are mandated to patch by February 13, 2026, as reported by CISA. <a href="https://app.daily.dev/posts/sI8icXzRd" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI agents are breaking traditional IAM security models</strong>: AI agents are fundamentally incompatible with traditional Identity and Access Management (IAM) systems, creating significant authorization risks. Unlike human users, agents often accumulate broad, persistent permissions without clear ownership or accountability, leading to potential bypasses and untraceable actions. Experts warn that ‘organizational agents’ shared across teams pose the greatest threat, necessitating new security protocols to track relationships between users, agents, and systems. <a href="https://app.daily.dev/posts/nUFRFzE1V" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Oracle’s Orasort patent expires, opening doors for open source databases</strong>: Oracle’s patent for the ‘Orasort’ sorting algorithm has expired after 20 years, making it available for open-source adoption. This algorithm, which combines common prefix skipping, adaptive quicksort/radix sort, and key substring caching, reportedly offers up to a 5x performance improvement. According to inventor Mark Callaghan, the open-source database community, particularly MySQL and PostgreSQL developers, are already considering integrating it to boost efficiency. <a href="https://app.daily.dev/posts/aheyA2T6F" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Good to know</h2>
<ul>
<li><strong>Microsoft Aspire expands language support for distributed apps</strong>: Microsoft’s Aspire platform now supports JavaScript, TypeScript, Python, and Java, broadening its appeal beyond .NET for distributed application development, especially for AI/ML and web projects. <a href="https://app.daily.dev/posts/gRAkh5lrI" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Skip tool open-sourced for cross-platform mobile development</strong>: Skip, a tool that converts Swift code for native iOS and Android apps using SwiftUI and Jetpack Compose, has been fully open-sourced, removing all paywalls and shifting to community funding via GitHub Sponsors. <a href="https://app.daily.dev/posts/7snXi8ahF" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Google Cloud offers public datasets via Apache Iceberg REST Catalog</strong>: Google Cloud now provides public datasets, like the NYC Taxi dataset, through the Apache Iceberg REST Catalog on BigLake, allowing read-only access via any Iceberg-compatible compute engine without infrastructure setup. <a href="https://app.daily.dev/posts/tn371FITl" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Compyle drastically reduces cloud terminal latency</strong>: Compyle has reduced cloud development environment terminal latency from 200ms to 14ms by implementing a warm pool, removing a socket server middleman, and deploying multi-regional sandbox pools closer to users. <a href="https://app.daily.dev/posts/2R6Ajv0Sw" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Newelle AI assistant for GNOME adds local LLM support</strong>: Newelle, the AI virtual assistant for the GNOME desktop, now supports local large language models (LLMs) via llama.cpp and includes a system command execution tool, offering users a choice between cloud-based and on-device AI inferencing. <a href="https://app.daily.dev/posts/Au9UNs8LJ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>DXVK-NVAPI 0.9.1 introduces DLSS ratio override and fixes</strong>: DXVK-NVAPI version 0.9.1 allows manual override of DLSS scaling ratios via environment variables and fixes startup crashes in several games, including ‘Middle-earth: Shadow of Mordor’ and ‘The Last of Us Part 2’. <a href="https://app.daily.dev/posts/6hOCF9MBP" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>Creative Tim UI launches open-source React/Tailwind components</strong>: Creative Tim UI has released a library of over 100 open-source UI components and blocks built on shadcn/ui, using React and Tailwind CSS, designed for e-commerce, dashboards, and Web 3.0 applications. <a href="https://app.daily.dev/feed-by-ids?id=1knteyTlI&amp;id=ArX5kf4A7" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
<h2>Bullshit police</h2>
<ul>
<li><strong>AI is making software engineering all about SRE, apparently</strong>: Two identical articles claim AI makes code-writing so easy that the new ‘core value’ for software engineers is operational excellence and reliability. So, building a working demo is ‘straightforward’ now, but maintaining it is the real challenge. Apparently, Site Reliability Engineering (SRE) is the future, because who needs to write good code when AI can just churn it out, right? It’s almost like they’re saying AI is great for quantity, but humans are still stuck with the quality control. <a href="https://app.daily.dev/posts/otuC5YedQ" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>New programming language ‘Coi’ claims to beat React and Vue</strong>: A new component-based language called Coi, targeting WebAssembly, is making bold claims about performance superiority over React and Vue. Their benchmarks show canvas rendering at 100 FPS versus Emscripten’s 40 FPS, achieved by ‘shared memory architecture’ and ‘O(1) reactivity.’ While it’s always fun to see new tech, these claims are based solely on Coi’s own publicity, so take that ‘beats React/Vue’ headline with a grain of salt until some independent third-party actually verifies it. <a href="https://app.daily.dev/posts/iOGOTjzat" target="_blank" rel="noopener nofollow">Read more</a></li>
<li><strong>AI coding agents: the energy footprint nobody talks about</strong>: Turns out, your AI coding agent isn’t just a token-guzzling genius; it’s also an energy hog. A median Claude Code session uses about 41 Wh, which is 138 times more than a typical chatbot query, according to Tech Lead Digest. Daily usage can hit 1,300 Wh, comparable to running an extra dishwasher cycle. So, while we’re all marveling at AI’s productivity, maybe we should also start thinking about its carbon footprint. Or, you know, just keep ignoring it like we do with everything else. <a href="https://app.daily.dev/posts/ex1Cm3GIg" target="_blank" rel="noopener nofollow">Read more</a></li>
</ul>
