SShortSingh.

Programming

0
ProgrammingDEV Community ·

Developer Uses Hackathon to Discover CDO Role Matching His Own Profile via RSS and AI

The HackAVP, described as New Caledonia's first hackathon dedicated to public sector job vacancies (AVPs) from OPT-NC, was launched this week. During a pitch session, a participant highlighted the availability of an RSS feed on the platform, noting its utility for quick, low-code integrations. As a practical demonstration on the SaaS track, he built a scheduled job within Claude CoWork to automatically find, match, and prepare job application documents stored directly on Google Drive. Without writing any code, the system identified vacancy 3134-26-1315_sr as the best match for his profile — which turned out to be a Chief Data Officer opportunity within his own organization. The demo was submitted via the hackathon's online form as a concrete example of what participants can build on the platform.

0
ProgrammingHacker News ·

AI Bioweapons Risk Report Sparks Debate Among Experts

A recent report warning about artificial intelligence's potential role in developing bioweapons has drawn sharply divided reactions from the scientific community. Some experts describe the findings as a serious and credible warning that warrants urgent attention. Others argue the report overstates the risks and may amount to an overreaction. The disagreement reflects a broader ongoing tension in the AI safety field over how to assess and communicate emerging threats. The debate highlights the challenge of balancing transparency about potential dangers with avoiding unnecessary alarm.

0
ProgrammingDEV Community ·

What AGI Actually Means and Why Today's AI Systems Fall Short of It

A technical analysis argues that current AI systems, including those from OpenAI, do not qualify as Artificial General Intelligence because they rely on deterministic, input-output processing rather than autonomous reasoning. Modern AI operates over stateless HTTP requests, requires users to resend full conversation history each time, and uses caching techniques that prevent genuine learning or self-correction. The author contends that so-called 'hallucinations' are simply a byproduct of next-word prediction, not evidence of understanding. True AGI, the piece suggests, would require persistent memory, continuous operation without request-response cycles, and dedicated hardware capable of growing autonomously. Until those conditions are met, today's AI is described as sophisticated autocomplete rather than genuine general intelligence.

0
ProgrammingDEV Community ·

Researchers Show Hidden LLM Reasoning Can Be Extracted in Two API Calls

A newly published research paper titled 'Stealing Reasoning Traces from Proprietary LLM APIs' reveals that encrypted chain-of-thought blocks returned by commercial LLMs like Claude, GPT-4, and Gemini can be exploited to expose hidden model reasoning. By replaying these blocks into a weaker model and jailbreaking it, researchers recovered internal reasoning steps using just two API calls. The technique bypasses provider safety layers without directly attacking the primary model, as demonstrated in a case involving a math-solving task. Organizations using LLMs with proprietary data face risks including intellectual property leakage, privacy regulation violations under frameworks like GDPR, and erosion of trust in AI-assisted workflows. Security experts recommend auditing prompt patterns, masking sensitive data before API submission, and selecting providers that actively strip chain-of-thought traces from responses.

0
ProgrammingDEV Community ·

Developer Finds Security Scanner Misses JSON Secrets After Building Deliberate Trick Tests

A developer tested their own MCP security scanner, mcpscan, using a purpose-built harness designed to probe each detection rule with synthetic attack samples. Six of seven test fixtures triggered findings, but the secrets detection rule failed entirely on a JSON config containing a live API key. Investigation revealed two flaws: the token pattern rejected keys with internal dashes, and a fallback regex broke on JSON's quoted key format — meaning the scanner silently missed the most common MCP config structure. A targeted patch fixed both issues, though two obfuscation-based bypasses involving string splitting and zero-width characters remain open and documented. The developer argues that a scanner aware of its own blind spots is more reliable than one claiming complete coverage without adversarial testing.

0
ProgrammingDEV Community ·

Engineer Deploys 4-Layer Security Stack on 29-Node RKE2 Cluster Before Any Workloads Run

A platform engineer has documented the security hardening of a 29-node RKE2 Kubernetes cluster spanning three distributed Proxmox datacenters, deliberately installing protections before any tenant workloads are deployed. The setup integrates four cloud-native tools: Kyverno for admission control, KubeArmor for kernel-level syscall enforcement, Falco for behavioral anomaly detection, and Trivy Operator for continuous CVE and misconfiguration scanning. Each layer operates independently, ensuring no single point of failure across the defensive stack. The entire configuration is managed through a fully GitOps-driven workflow via ArgoCD, replacing all ad-hoc terminal commands with version-controlled commits. This work forms the third episode in a broader infrastructure series covering Terraform provisioning, HA RKE2 cluster setup, and now platform-wide security hardening.

0
ProgrammingHacker News ·

2011 Essay Warns of Threats to General-Purpose Computing Resurfacing Online

A 2011 essay titled 'The Coming War on General Computation' has resurfaced on Hacker News, drawing renewed attention. Originally delivered as a talk, the piece argues that governments and industries may increasingly seek to restrict general-purpose computers in favor of locked-down, single-purpose devices. The author contends that such restrictions, often framed around copyright or security concerns, pose a fundamental threat to user freedoms and open computing. The post garnered 20 points on Hacker News, sparking a small but notable discussion among readers.

0
ProgrammingDEV Community ·

Blazor WebAssembly App Runs Semantic Search on GitHub Pages With No Backend

A developer has implemented semantic search on the Blazing Story documentation site, a Blazor WebAssembly standalone app hosted entirely on GitHub Pages with no server, API, or database. The approach splits work into two stages: at build time, a C# console app generates vector embeddings from Markdown files and saves them into a static index file; at runtime, the browser loads that index and computes embeddings only for the user's search query. Vector search works by converting text into numerical arrays and measuring cosine similarity to find content that is close in meaning, even when exact keywords do not match. This allows users to search documentation semantically — for instance, querying 'fruit' could surface pages mentioning 'apple' or 'banana' — without any client-server round trips. The solution relies on Microsoft's ML.NET ONNX runtime and tokenizer NuGet packages to run a natural language processing model directly inside the browser via WebAssembly.

0
ProgrammingDEV Community ·

DeepSeek V4.1-Flash Cache Pricing at $0.003 Reshapes AI Model Cost Comparisons

A September 2026 analysis highlights that cache hit pricing — not just per-token cost — is a critical but widely overlooked factor when comparing AI models for agent workloads. DeepSeek's V4.1-Flash offers cache hits at $0.003 per million tokens during off-peak hours, compared to $0.30–$0.50 for rivals like Kimi K3, GPT-5.6 Sol, and Claude Opus 5, a difference of up to 167 times. Because agent tasks repeatedly read the same context — such as repositories, documents, and conversation history — cache costs can dominate total expenditure, with a sample 50-million-token cache scenario costing roughly $0.15 on V4.1-Flash versus $25 on Claude Opus 5. On September 13, 2026, Together AI announced V4.1-Flash availability on its platform, citing benchmark results from DeepSeek's own model card showing the model outperforming GPT-5.6 Sol on several agent-focused tests. A July 2026 VentureBeat survey of 170 organizations found that 53% do not rigorously track AI costs, suggesting much of the market may be unaware of the pricing gap that cache hit rates create.

0
ProgrammingDEV Community ·

AI Makes Prototypes Cheap, But Ownership Still Determines Long-Term Success

AI tools have dramatically reduced the time and cost to build early-stage prototypes and MVPs, but the challenges of reliability, governance, and adoption remain as expensive as ever. Organizations that continue funding short-term projects tend to accumulate unowned technical debt once teams disband after launch, leaving the same underlying problems unresolved. Experts including Marty Cagan and researchers cited in Harvard Business Review argue that permanent, outcome-accountable product teams consistently outperform temporary project-based structures. Frameworks like Team Topologies advocate for long-lived, stream-aligned teams with clear ownership and end-to-end accountability over milestone-driven handoffs. The core shift required is moving from asking when something ships to asking who is responsible for improving it next quarter.

0
ProgrammingDEV Community ·

DeepSeek V4.1-Flash cuts KV cache 4x despite doubling model size, targeting AI agents

DeepSeek released its V4.1-Flash model on September 10, 2026, specifically engineered to reduce the cost of running AI agents that repeatedly read large contexts rather than generate long outputs. The model introduces a Causal Encoder-Decoder (CED) architecture that splits its 552 billion parameters into encoder and decoder halves, sharply reducing active parameters during input processing to just 8 billion. Additional techniques — including Compressed Sparse Attention 2 (CSA2) and FP4 KV caching — bring total KV cache storage down to 890 bytes per token, roughly one-quarter of its predecessor V4-Flash and 437 times smaller than the original V1. This matters economically because agentic workloads spend most compute re-reading the same context in loops, meaning KV cache — not token generation — is the dominant cost driver. Benchmark scores for agent tasks such as DeepSWE and Terminal-Bench rose substantially, from 54.4 to 74.2 and 82.7 to 90.6 respectively, compared to V4-Flash.

0
ProgrammingDEV Community ·

Coolify, Dokku, CapRover or Ownkube: Which Self-Hosted PaaS Fits Your Team in 2026?

As teams look to cut Heroku costs in 2026, four self-hosted PaaS platforms — Coolify, Dokku, CapRover, and Ownkube — have emerged as the most commonly evaluated alternatives. Dokku and CapRover are best suited for solo developers or indie projects running on a single low-cost VPS, with Dokku offering Heroku-like buildpack support and CapRover providing a click-through UI. Coolify targets solo founders and small teams seeking a polished interface with built-in support for databases, queues, and monitoring. Ownkube is positioned for teams of 5–20 engineers on AWS who want Kubernetes-grade scalability without hiring dedicated DevOps staff, operating the platform on the user's own cloud account with a managed control plane. The key tradeoff across all four is operational burden — while each delivers Heroku-style deployment primitives, only Ownkube offloads ongoing platform maintenance, whereas the others leave infrastructure operations largely to the user.

0
ProgrammingDEV Community ·

AWS NAT Gateway Can Cost More Than Your Entire EC2 Bill — Here's the Fix

AWS NAT gateways charge $0.045 per hour per availability zone plus $0.045 per GB of data processed, meaning a multi-AZ production workload pushing 10 TB per month can rack up over $1,400 in NAT costs alone — often exceeding the compute bill. The core problem is that traffic between private subnets and AWS services like S3, ECR, and Secrets Manager is routed through the NAT gateway by default, generating unnecessary data-processing fees. The most impactful fix is enabling VPC endpoints: gateway endpoints for S3 and DynamoDB are free, while interface endpoints for other AWS services charge $0.01 per GB instead of $0.045, breaking even at roughly 700 GB per month. For outbound IPv6 traffic, AWS's egress-only internet gateway is free and carries no data-processing or hourly fees, yet remains widely underutilized. Teams are advised to audit which traffic actually needs NAT and reroute internal AWS service calls through endpoints as a first step to significantly reducing cloud costs.

0
ProgrammingDEV Community ·

AI Avatar v20 Adds Voxel Garden, Cursor Pop-ups, and AI Tool Notifications

AI Avatar, a free VS Code and Chrome extension that animates a VRM avatar during coding sessions, has released version 20 with several new features. The update introduces a Minecraft-style voxel flower garden that grows automatically when triggered by Claude Code or manually via clicks, complete with butterflies, rain, and rainbows. A new Cursor Avatar spin-off Electron app lets the avatar appear anywhere on the desktop outside the browser or editor side panel. The release also includes Notification Avatar, which uses hooks to alert users when AI coding tools like Claude Code or Codex stop or require approval. Additional updates include a Kaleidoscope Mode for randomized backgrounds and support for Piper TTS as a simpler text-to-speech alternative.

0
ProgrammingDEV Community ·

AI Agents Are Moving From Answering Questions to Taking Action

A growing shift in AI usage is moving beyond simple prompt-and-response interactions toward autonomous agents that can plan, use tools, and execute multi-step workflows with minimal human input. When equipped with web search, terminal access, APIs, and persistent memory, AI systems become capable of acting on goals rather than just generating text. Demonstrations like the Grok Bot Galaxy showcase how AI could handle complex tasks — such as diagnosing a conversion drop or preparing a marketing campaign — by chaining individual steps together. However, this increased autonomy raises serious questions about permissions, oversight, and what happens when an AI misinterprets its goal, especially when it has access to production systems, customer data, or payment platforms. The key design challenge is no longer whether AI can complete a task, but how much of a workflow should be delegated to it and what safeguards must be built around that autonomy.

0
ProgrammingDEV Community ·

Developer Catches SAP OData API Silently Capping Data at 1,000 Rows

A developer querying SAP Internal Orders via an OData API received exactly 1,000 rows — a suspiciously round number that prompted further investigation. Standard OData pagination parameters such as $skip and $top were ignored by SAP's gateway, which continued returning the same first 1,000 records regardless. Even a $count request over the full date range returned 1,000, indicating the cap affected metadata as well as data. Splitting the query into two date-range halves revealed a true total of 1,589 rows, meaning 589 records had been silently dropped with no error or warning. The incident highlights the need for independent verification strategies — such as windowed queries and cross-checked partial counts — rather than relying solely on an API's built-in pagination mechanisms.

0
ProgrammingDEV Community ·

Matching Prediction Markets Across Venues Is Harder Than It Looks

A developer building a cross-venue price comparison tool for prediction markets found that the biggest challenge was not aggregating order books but determining whether two markets on different platforms refer to the same event. Platforms like Polymarket and Kalshi use entirely different naming conventions, slugs, and ticker formats, making automated matching unreliable — for example, a fuzzy title match on a Fed rate decision can accidentally pair five separate Fed meeting markets. Even within a single venue, the same question can appear as two distinct listings with unrelated identifiers, falsely signaling arbitrage opportunities to any matcher that assumes internal uniqueness. The developer identified resolution date, resolution source, outcome buckets, and settlement rules as the most discriminating factors for accurate matching, with price comparison coming last. Both Polymarket and Kalshi offer read-only public APIs without requiring authentication, and the developer published a full reference guide along with a live matching terminal called Preduck covering data from September 13, 2026.

0
ProgrammingDEV Community ·

Developer Builds End-to-End CI/CD Pipeline Using Azure DevOps, AKS, Helm and PostgreSQL

A developer has documented the process of integrating Azure DevOps, Azure Kubernetes Service, Helm, and PostgreSQL into a single production-style CI/CD pipeline after learning each technology separately. The project uses a simple three-tier application — frontend, backend, and database — to focus learning on infrastructure rather than application complexity. Code moves from a GitHub repository through Azure DevOps pipelines, gets containerized and pushed to Azure Container Registry, and is then deployed to AKS via Helm charts. PostgreSQL was deliberately hosted as an Azure Flexible Server outside the cluster, connected via a private network and private DNS zone, rather than running inside Kubernetes. The project, built on an Azure for Students subscription in the South Africa North region, is publicly available on GitHub and currently covers the DEV environment, with UAT and PROD stages planned for future iterations.

0
ProgrammingDEV Community ·

JQuickCurl Lets Developers Toggle curl Command Parts at Runtime Using XML

JQuickCurl, a Java-based HTTP client library, introduces conditional rendering for curl commands through an XML-based configuration mode. Developers can wrap curl fragments in conditional XML elements that are included or excluded based on runtime expressions, eliminating the need for if/else logic in application code. The feature supports toggling headers, logging flags, and request body fragments depending on parameters such as environment type, debug mode, or the presence of a trace ID. A runnable demo using httpbin.org illustrates how the same interface method can produce different wire requests based on boolean flags passed at call time. The approach aims to keep HTTP request definitions clean and environment-aware without sending empty or unintended headers in production.

0
ProgrammingDEV Community ·

Freelance Designer Learns Hard Lesson: Client Management Matters as Much as Skill

A freelance designer recently came close to abandoning a project not due to lack of ability, but because of persistent communication breakdowns with a client. Despite following the brief, the designer repeatedly received feedback that the work did not match expectations, with requirements often shifting between revisions. The experience highlighted that technical design skills alone are insufficient for freelance success — clear documentation, boundary-setting, and asking the right questions are equally critical. The designer concluded that knowing when a working relationship is no longer viable is itself an important professional skill. The difficult project ultimately offered lessons that smoother engagements rarely provide.

← NewerPage 1359 of 3312Older →