After rewriting SQLite in Rust, Turso turns its sights on Postgres

theregister.com Jul 30, 2026

After rewriting SQLite in Rust, Turso turns its sights on Postgres

Curated by Shubham · TechCapsules · Jul 30, 2026

AI-summarised brief · reviewed before publication

Cloud database provider Turso has launched a Postgres-compatible implementation built upon its existing Rust-based foundation. Following the successful rewrite of SQLite, originally codenamed Limbo, the company aims to expand its virtual machine architecture to support multiple SQL frontends. Turso CEO Glauber Costa stated that while Postgres is fundamentally sound, it requires modernization for the cloud-native era. The new initiative allows the same core to power different database systems, with MySQL and Redis frontends potentially following. Co-founders Glauber Costa and Pekka Enberg previously left ScyllaDB to create ChiselStrike, which pivoted to become Turso. They initially forked SQLite as libSQL to enable cloud features and external contributions. However, deeper architectural changes necessitated a complete rewrite. This new engine, named after the mythical Finnish sea creature Iku-Turso, serves as the planned foundation for future database engines. The company compares its approach to LLVM, offering one reliable core with various compiled frontends. This move marks a significant shift from simple forking to comprehensive re-engineering of established database technologies for modern cloud infrastructure requirements.

💡 Why It Matters

· By treating database engines as interchangeable frontends on a single modern core, Turso challenges the industry standard of maintaining separate, legacy codebases for each SQL dialect.
· This architecture promises reduced maintenance overhead and faster innovation cycles for developers tired of patching decades-old systems.

Read full article on theregister.com

Stadia was some of the most fun I’ve had gaming, but it…

Google’s Stadia cloud gaming service has been discontinued for nearly four years, a move the author deems correct despite the platform’s personal appeal. Stadia allowed high-end game streaming to various devices with minimal hardware, relying on robust internet connections to mitigate latency issues that previously hindered competitors like OnLive. While GeForce Now and console streaming services persist, Stadia offered a frictionless experience without complex account linking or launchers. The author enjoyed playing titles like Red Dead Redemption 2 and Cyberpunk 2077 across different locations, praising the seamless switching between desktop and mobile. However, Stadia suffered from a limited game library and a disastrous 2019 GDC launch where Google overpromised capabilities. Network restrictions and initial performance issues alienated many users. Although the platform provided enjoyable casual gaming experiences, it failed to capitalize on the pandemic-driven demand for home entertainment. The author suggests that even if Stadia had survived, emerging AI technologies and shifting gaming landscapes would likely have rendered it obsolete, confirming its discontinuation was inevitable.

💡 Stadia’s failure underscores that superior user experience cannot overcome fundamental market missteps like poor launch execution and limited content. Its demise validates that cloud gaming success depends more on library depth and infrastructure reliability than just technical convenience.

siliconangle.com · Jul 31

Amazon’s stock pops on roaring cloud growth and soaring AI demand

Amazon.com Inc. reported second-quarter earnings of $5.75 per share, significantly exceeding Wall Street’s $1.82 forecast, driven by a $53 billion non-operational gain from its Anthropic stake. Total revenue reached $200.61 billion, a 20% year-over-year increase, while net income surged to $62.6 billion. Amazon Web Services delivered $42.2 billion in revenue, marking a 37% increase and the unit’s fastest growth since 2021. CEO Andy Jassy attributed this acceleration to surging AI demand and the success of homegrown Trainium and Graviton chips, which now generate a $25 billion annual run rate. These custom processors serve as alternatives to Nvidia’s GPUs, helping diversify the cloud business. Investors responded positively to the strong cloud performance and AI momentum, pushing Amazon’s stock up more than 9% in late trading. The results demonstrate Amazon’s ability to capitalize on enterprise AI development efforts through platforms like Amazon Bedrock, reinforcing its competitive position against rivals Microsoft and Alphabet in the public cloud infrastructure market.

💡 Amazon’s success with custom silicon proves hyperscalers can reduce reliance on Nvidia, fundamentally shifting supply chain power dynamics in the AI hardware market.

nationthailand.com · Jul 31

Edge AI: Why It Could Be the Next Strategic Frontier for Thai…

Edge AI is emerging as a strategic frontier for Thai enterprises, offering a flexible alternative to cloud-dependent systems. While cloud AI powers the current boom, it introduces latency, API costs, and data sovereignty concerns. Open-weight models from companies like Google and Meta now allow sophisticated AI to run directly on local devices, including laptops and industrial equipment. This shift enables a hybrid approach, balancing cloud compute with local deployment for speed, privacy, and resilience. Tools like Ollama facilitate easy local deployment, already adopted by many Fortune 500 firms. For Thailand, this technology lowers barriers to entry, allowing organizations to build models tailored to local languages and regulations without training frontier models from scratch. The Typhoon family of Thai language models, built on open-weight foundations, exemplifies this potential. By leveraging existing global models, Thai enterprises can address specific local needs efficiently, contributing to the global open-model community while maintaining operational control and data security.

💡 Localized AI deployment empowers Thai businesses to bypass reliance on foreign cloud providers, securing data sovereignty and reducing operational costs. This technological shift transforms Thailand from a passive consumer of global AI into an active contributor to the open-model ecosystem.

appleinsider.com · Jul 31

From iTools to iCloud: Apple’s first foray into Software as a Service

Apple’s iCloud, now central to its ecosystem, evolved from three predecessor services launched over a decade ago. In 2000, Apple introduced iTools, a free suite for Mac OS 9 users featuring iCards, KidSafe, HomePage, iDisk, and mac.com email. iReview and KidSafe were discontinued in 2001. iDisk offered 20MB of free storage, with optional paid upgrades. In July 2002, Apple rebranded iTools to .Mac, introducing a subscription fee. .Mac added Backup capabilities and McAfee antivirus. By 2007, it offered 10GB of shared storage for email and iDisk. In 2008, Apple launched MobileMe after acquiring me.com. CEO Steve Jobs positioned it as a consumer alternative to Microsoft Exchange, enabling push email, calendar, and contact synchronization across platforms. MobileMe allowed web access and cross-platform syncing, laying the foundational infrastructure for the deep ecosystem integration that defines modern Apple services and the eventual launch of iCloud in late 2011.

💡 This history reveals Apple’s strategic pivot from free utilities to paid subscription models, establishing the technical and commercial blueprint for its current services revenue. Understanding this evolution clarifies how early experiments with storage and sync directly enabled the seamless cross-device integration that now locks users into the Apple ecosystem.

infosecurity-magazine.com · Jul 30

The Tesla Hack is a Serious Cryptojacking Warning

Tesla’s Amazon Web Services cloud infrastructure was recently compromised by hackers who deployed malware to illegally mine cryptocurrency. This incident highlights the rapid emergence of cryptojacking as a distinct cyberattack class, a category barely recognized just one year ago. The breach occurred due to insecurely configured Kubernetes clusters within Tesla’s AWS environment. Although Tesla responded swiftly to remove the malware and secure its systems, minimizing fallout, the event underscores a growing threat to businesses and critical infrastructure. Cybercriminals are increasingly targeting large organizations not only for data theft but also to hijack computing resources for profit. With 63% of Kubernetes stacks running on AWS, the widespread adoption of this container technology, combined with management complexities and configuration errors, creates significant vulnerabilities. Enterprises face risks beyond inflated cloud bills, including potential compromise of sensitive keys and data. The attack demonstrates that sophisticated techniques are evolving, necessitating robust security measures like blind spot detection, continuous monitoring, and automated remediation to protect against these emerging threats.

💡 The breach exposes how widespread reliance on complex container technologies creates systemic vulnerabilities across the tech sector. It forces enterprises to prioritize rigorous configuration management over mere productivity gains to prevent resource hijacking.

prtimes.jp · Jul 29

Grafana Labs、「2026年版 Gartner(R) Magic Quadrant(TM) For Observability Platforms」でリーダーに選出

Grafana Labs has been named a Leader in the 2026 Gartner Magic Quadrant for Observability Platforms, marking its third consecutive year in this position. The company also achieved the highest ranking for Completeness of Vision for the second year running. This recognition highlights Grafana’s focus on open, composable observability tailored for complex, AI-driven systems. According to Grafana’s 2026 survey, operational complexity remains the primary challenge, though observability adoption has risen to half of all organizations. To address this, Grafana Cloud now features built-in AI tools like Grafana Assistant and native agent access via CLI and MCP servers. These innovations enable teams to monitor LLMs, agents, and inference pipelines alongside traditional infrastructure. Additionally, Adaptive Telemetry helps optimize costs by reducing noise. In Gartner’s Critical Capabilities report, Grafana Cloud scored 4.4 out of 5 for AI/LLM Observability. The company emphasizes that open standards like OpenTelemetry are essential for empowering both human engineers and AI agents in modern, distributed environments.

💡 As AI agents become core infrastructure, observability must evolve beyond human-centric dashboards to support machine-speed diagnostics. Grafana’s leadership validates the shift toward open, agent-native platforms that prevent data silos and enable autonomous system management.