Beyond Market Intelligence/Steef-Jan Wiggers

Steef-Jan Wiggers

Steef-Jan Wiggers at Beyond Market Intelligence is a file of 24 stories. The newest of them: “Explore how PrivateLink now opens entire network segments across AWS accounts”, “AWS Nitro Sandboxes Isolate Each EKS Pod in Its Own MicroVM at No Extra Cost”, and “Google Cloud locks API stability with predictable quarterly version snapshots”. AWS has detailed tunnel endpoints, a PrivateLink type that opens entire network segments across accounts rather than one resource at a time. AWS Nitro Sandboxes give each opted-in EKS pod its own Firecracker microVM with a separate kernel, at no extra cost. Acme AI is the next-generation, AI-powered spreadsheet platform built to replace Excel and redefine how analysts, data scientists, and enterprise teams work… The list below is every Steef-Jan Wiggers story on Beyond Market Intelligence, newest first.

Explore how PrivateLink now opens entire network segments across AWS accounts
InfoQ

Explore how PrivateLink now opens entire network segments across AWS accounts

AWS has detailed tunnel endpoints, a PrivateLink type that opens entire network segments across accounts rather than one resource at a time. This matters because it solves overlapping address spaces and relies on consumer-initiated connections only, but the trade-offs are real: traffic must be wrapped in GENEVE, only TCP and DNS pass, and access logs can't trace traffic to a specific workload. It's a powerful, if opinionated, step toward simpler cross-account networking. For more on isolating workloads, see our coverage of AWS Nitro Sandboxes.

AWS Nitro Sandboxes Isolate Each EKS Pod in Its Own MicroVM at No Extra Cost
InfoQ

AWS Nitro Sandboxes Isolate Each EKS Pod in Its Own MicroVM at No Extra Cost

AWS Nitro Sandboxes give each opted-in EKS pod its own Firecracker microVM with a separate kernel, at no extra cost. A single `runtimeClassName` field enables this isolation, while a confidential mode adds attestation-gated key release. The trade-offs are honest: no GPU passthrough, a five-sandbox-per-vCPU cap, and half the compute on SMT instances. This is practical, not flashy. For more on infrastructure control, see our coverage of Google Cloud's predictable API versioning.

Google Cloud locks API stability with predictable quarterly version snapshots
InfoQ

Google Cloud locks API stability with predictable quarterly version snapshots

Google Cloud has locked API stability into a predictable rhythm. The Compute Engine API now offers quarterly, immutable version snapshots like 2026-09-01, selected per request via header or parameter. This replaces the old single mutable /v1 surface with a model Azure Resource Manager and Stripe have used for years. New versions can adjust defaults, error formats, or IAM checks without touching /v1 callers. It is a practical, progressive move that lets teams explore updates on their own terms.

Explore how Azure Logic Apps CLI brings deterministic control to AI-assisted development
InfoQ

Explore how Azure Logic Apps CLI brings deterministic control to AI-assisted development

Microsoft's public preview of the logicapp extension for the Azure CLI brings deterministic control to AI-assisted development by moving workflow authoring and test scaffolding into the terminal. This is a smart shift for teams who want precision without the overhead of a full VS Code environment. If you're exploring how local AI tools need grounding, our piece on running local LLMs with Apple's fm CLI offers a useful parallel.

Explore how Google Cloud spend caps now pause services automatically each month
InfoQ

Explore how Google Cloud spend caps now pause services automatically each month

Google Cloud's spend caps now automatically pause services each month once a budget is hit, without deleting your resources. The generally available feature adds an API, estimated live costs, and automatic monthly resets. It's a practical step toward cost control, though it currently covers only the Gemini API, Agent Platform, Cloud Run, and Cloud Run functions, and overages during enforcement delays still get billed. For more on taking control of your infrastructure, see our coverage of Google Cloud's self-managed VMware solution.

Google Cloud hands VMware lifecycle control back to you on bare metal
InfoQ

Google Cloud hands VMware lifecycle control back to you on bare metal

Google Cloud now offers a self-managed VMware option on bare metal, giving you root access across ESXi, vCenter, and NSX in 22 regions. It shifts lifecycle control, patching, and Broadcom licensing back to you. This sits beside the managed VMware Engine and matches what AWS and Oracle provide. For teams who want full command of their infrastructure without losing VMware's power, this is a welcome move. Explore how GitLab 19.5 brings governed AI agents into your workflow for another take on platform control.

AWS Debuts Continuum to Automate Code Fixes and Prove Exploitability
InfoQ

AWS Debuts Continuum to Automate Code Fixes and Prove Exploitability

AWS has made Continuum generally available in two regions, and it does something most vulnerability tools only promise: it proves exploitability by running real exploits in a sandbox before opening fixes as pull requests. Pricing is per finding and per line of code, which raises a fair question, what would justify letting it merge fixes on its own? As autonomous AI agents gain more power, the guardrails conversation becomes urgent. For deeper context, see our coverage of Docker's guide to AI agent security risks.

AWS Launches ECS Gateway with Auto-Generated Hostnames for Simpler Networking
InfoQ

AWS Launches ECS Gateway with Auto-Generated Hostnames for Simpler Networking

Platform teams at AWS now have a cleaner path to simplified networking. ECS Gateway lets them define ingress once, while application teams opt in with a connector. Each service automatically receives a generated hostname with host-based routing. It's a focused, practical step forward, though custom domains remain absent for now. For teams managing large-scale data movement alongside container workloads, our coverage of AWS File Mover Express offers a complementary look at reducing operational friction.

Cloudflare's post-quantum key guard enters closed beta with a fail-closed design
InfoQ

Cloudflare's post-quantum key guard enters closed beta with a fail-closed design

Cloudflare's post-quantum rebuild of GeoKey Manager is now in closed beta, and its fail-closed design is the detail that demands attention. Keys live only in kernel memory on verified regional servers; any failed check rejects the request outright. That approach prioritizes security over availability, which practitioners warn could trigger a regional outage. General availability arrives Q1 2027. For more on how infrastructure security is evolving, see our coverage of Fortanix tying Kubernetes key release to workload attestation.

LocalStack's Azure Emulator Stores Your Security Rules but Won't Enforce Them
InfoQ

LocalStack's Azure Emulator Stores Your Security Rules but Won't Enforce Them

LocalStack's Azure emulator now runs your infrastructure-as-code locally, ARM, Bicep, and Terraform included, without rewrites. That's genuinely useful. But here's the catch: the emulator stores security rules without enforcing them. Access restrictions, TLS, and network rules are accepted and returned, yet never applied. Developers should treat this as a validation tool, not a security sandbox. For deeper context on cloud-native security patterns, explore how Fortanix ties Kubernetes key release to workload attestation.

Cloudflare unlocks Quicksilver speed for Workers KV in private beta
InfoQ

Cloudflare unlocks Quicksilver speed for Workers KV in private beta

A single read request in Cloudflare's classic Workers KV takes 287 milliseconds at p99. That same operation in the new private beta, which taps the internal Quicksilver key-value store, finishes in 1.62 milliseconds. The tradeoffs are real: 10,000 keys per namespace, one megabyte per namespace, one write per second. This isn't for everything, but for the workloads it fits, the speed leap is dramatic. For more on Cloudflare's performance bets, see our coverage of its AI Gateway's Auto Router.

AI time savings drop sharply as developer friction returns to focus
InfoQ

AI time savings drop sharply as developer friction returns to focus

Only 26% of developers now say AI saves them ten or more hours a week, down from 68% a year ago, according to Atlassian's latest State of Developer Experience report. The friction AI was supposed to eliminate is creeping back, and engineering leaders are spending more time managing AI output than their teams do. The tool isn't the shortcut we hoped for. For a different take on developer productivity, see our coverage of DigitalOcean's new data services for building smarter AI agents.

Cloudflare's AI Gateway Picks the Right Model for Each Request Automatically
InfoQ

Cloudflare's AI Gateway Picks the Right Model for Each Request Automatically

Cloudflare's Auto Router, now in public beta through AI Gateway, automatically selects the right model for each request by scoring complexity, ambiguity, stakes, and context dependence. Cloudflare's own benchmark places it ten points below Claude Opus 5.5 on success rate at a third of the cost. That efficiency is compelling, though practitioners rightly question how stakes hold up across long agent sessions.

PHP Patches Two Critical Flaws Four Days After Attackers Struck
InfoQ

PHP Patches Two Critical Flaws Four Days After Attackers Struck

Four days after an attacker exploited a critical PHP vulnerability in the wild, the project shipped patches. That gap is uncomfortable. PHP versions 8.5.12, 8.4.27, 8.3.36, and 8.2.35 fix two flaws allowing unauthenticated remote code execution. Operators scrambled just to determine whether they ran PHP, where to get the fix, and how to confirm it landed. Speed matters, but so does a clear update path. For a deeper look at how teams navigate such moments, see our related piece on PingCAP's recent cyberattack.

Open-Source Adobe Clones Hit 45% Parity, Still Months from Ready
InfoQ

Open-Source Adobe Clones Hit 45% Parity, Still Months from Ready

A developer used Claude Opus 5.5 to build open-source replacements for seven Adobe applications, promoting them on Reddit with a projection of near-complete parity within a month. Then told Hacker News the project was not close to ready. Its own roadmap puts it at 45% readiness for real work, requiring 2,300 to 4,100 agent-hours to reach full parity. Ambitious vision, but honesty about the gap earns trust. For deeper coverage, explore how dotnetup simplifies .NET management without admin rights.

AWS Tests European Cloud Independence Without Global Network Link
InfoQ

AWS Tests European Cloud Independence Without Global Network Link

On October 24, AWS will sever its European Sovereign Cloud from the global network backbone, running on EU hardware, software, and local staff. The company insists neither system is operation-critical, yet practitioners remain split on whether this test actually addresses customer concerns. It is a deliberate exercise in independence, but the question lingers: does isolation prove sovereignty? For deeper context on shifting AI alliances, see our coverage of Replit pivoting from Anthropic to OpenAI.

Microsoft's Maia 300 Targets Multi-Gigawatt Inference While Efficiency Claims Face Scrutiny
InfoQ

Microsoft's Maia 300 Targets Multi-Gigawatt Inference While Efficiency Claims Face Scrutiny

Microsoft's plan to scale Maia to multi-gigawatt deployments by 2027 signals serious intent, but leading the announcement with tokens-per-watt invites scrutiny. Practitioners are right to question that efficiency claim. Maia 300's design is clever, collapsing scale-up and scale-out onto one Ethernet fabric and moving KV state between prefill and decode workers without leaving it. For teams focused on shipping faster, our piece on how AI helps developers accelerate release cadence offers a complementary perspective on the productivity side of the equation.

Cloudflare's new gateway keeps decryption separate from its own traffic
InfoQ

Cloudflare's new gateway keeps decryption separate from its own traffic

Cloudflare's OHTTP Gateway is drawing attention for what it refuses to do: decrypt traffic arriving from its own Workers or proxied hosts. That separation matters, because it keeps client identifiers and request contents out of the same hands. It's a thoughtful constraint in a privacy landscape where trust is often the first casualty. For deeper platform context, our coverage of Cloudflare welcoming Deno's expertise offers useful background on how the company is evolving its developer stack.

Cloudflare's K2 Beta Turns Object Storage Into a One-Second Event Stream
InfoQ

Cloudflare's K2 Beta Turns Object Storage Into a One-Second Event Stream

Cloudflare's K2 beta turns R2 object storage into a serverless event stream with roughly one-second produce latency at p99. That's a compelling shift for teams already invested in Cloudflare's ecosystem, though the 7.5-second end-to-end p99 reported by one tester suggests the full path still has room to tighten. Consume charges matching produce charges keeps pricing predictable. For deeper context on Cloudflare's AI ambitions, Clef offers a sharper decision-making model worth exploring.

Google's Spanner Omni goes live, swapping hardware clocks and storage for software
InfoQ

Google's Spanner Omni goes live, swapping hardware clocks and storage for software

Google made Spanner Omni generally available, swapping its own hardware clocks and storage for software abstractions. Running its distributed SQL database on-premises, across clouds, or on a laptop is a striking move, but the lack of an availability SLA means practitioners are pricing the trade-off in tail latency and operational toil. For more on how cloud services manage reliability under pressure, see our coverage of Cloudflare's cross-tenant data exposure vulnerability.

Cloudflare resolves data leak risk across container boundaries
InfoQ

Cloudflare resolves data leak risk across container boundaries

Recovering complete SQLite databases from a neighbor's container isn't a theoretical risk, it was a real vulnerability in Cloudflare's infrastructure. The company disclosed a cross-tenant data exposure flaw in Containers and Sandboxes, where thin-provisioned storage pools failed to zero reused blocks. Researchers pulled directory structures and database pages across four continents. Cloudflare fixed the issue and found no signs of exploitation. For more on how Cloudflare thinks about future-proofing security, explore their work on a quantum-safe path for internet security.

Python Workers Go Live as Questions on Speed and Upstream Support Emerge
InfoQ

Python Workers Go Live as Questions on Speed and Upstream Support Emerge

Cloudflare's Python Workers have gone GA, but the launch feels more like an opening act than a headliner. Built on PEP 783 and socket syscalls over the Workers connect API, the announcement conspicuously omits performance figures. That silence is loud. A urllib3 maintainer is already asking who supports the upstream work afterward, and Wasmer's founder wants cold-start numbers, citing about 1.027 seconds from Cloudflare's own blog. These are fair, pointed questions. If you're evaluating this for production, you'll want answers before committing.

Azure Repos code reviews now billed per review with five-concurrency limit
InfoQ

Azure Repos code reviews now billed per review with five-concurrency limit

Microsoft has opened GitHub Copilot code review for Azure Repos to all Azure DevOps customers, acknowledging that many teams aren't ready to leave for GitHub. That's a practical move, one that meets users where they are. Reviews bill per use through your linked Azure subscription, with costs showing up in Cost Management two days later. Budgets alert you but won't halt reviews, and concurrency caps at five per organization. It's a straightforward, usage-based approach.

Trace AI agents turn by turn with Cloudflare's new tracing spans
InfoQ

Trace AI agents turn by turn with Cloudflare's new tracing spans

Cloudflare's new agent tracing brings welcome visibility to AI-driven workflows, adding spans for agent invocations, model calls, tool runs, and approvals directly into existing Workers traces. Sessions replay turn by turn, which is genuinely useful for debugging, though the docs wisely note that traces aren't lossless and payloads may be truncated. Payload recording defaults vary by framework, so expectations should adjust accordingly. Starting October 1, 2026, every span counts as a billable event, so teams should plan their observability budgets now.