India rejects UN panel’s observations on its ‘rights violations’

Developer Launches Free AI Model Tracker Covering 900 Models and 2,300 Static Pages

A systems engineer has publicly launched theknowngood.com, a website that aggregates evaluation data for around 900 AI models, including benchmark results, pricing, arena ratings, and performance metrics. The site is built as fully static HTML rendered via Python and Jinja2 from a Postgres database, meaning no database queries occur at request time. All CSV and JSON data exports are freely available under a CC BY 4.0 license, with no user accounts, tracking, or inbound API. For security, nginx is bound to loopback only and the site is served through an outbound tunnel, eliminating public-facing open ports entirely. The project is currently self-hosted on home hardware, and the creator is actively seeking user feedback on performance and test coverage gaps.

LLM Cost Routing Cuts Inference Spend 48x and Fixes Broken Margin Curves

A software team discovered that one frontier AI model was handling 77% of their calls and consuming 97% of their inference budget simply because it was the default setting, not because most tasks required it. After measuring a week of production traffic, they found their average call cost $0.00524 on the frontier model versus $0.00011 on a capable open-weight alternative — a roughly 48x difference. Under a $20/month subscription plan, routing all calls through the frontier model turned high-engagement users into a financial loss, while task-based routing kept margins above 97% even at 5,000 messages per month. The key insight is that smart routing does not merely reduce average costs — it reverses the marginal cost curve so that heavier product usage compounds profit rather than loss. The team cautioned, however, that a cheaper model requiring multiple retries can quickly erase those savings, making reliable routing logic essential.

‘Qualifying standards cannot be diluted’: Government on screening test for foreign medical graduates

Managed vs Self-Hosted React Native OTA: It's About Ownership, Not Infrastructure

When choosing between managed and self-hosted Over-the-Air (OTA) update systems for React Native apps, the core difference lies in who owns operational responsibility, not where files are stored. Both approaches support channels, rollouts, rollbacks, and CDN delivery, but a managed service shifts platform operations to a provider while self-hosting places that burden on the organization. One responsibility that cannot be outsourced in either model is verifying that an OTA update is compatible with the native binary already installed on a user's device, since OTA can only replace JavaScript and bundled assets, not native modules or compiled capabilities. Self-hosted systems offer greater flexibility and control but require the organization to handle server upgrades, security patching, backups, and multi-generation client protocol compatibility. The right choice depends on an organization's size, regulatory requirements, and existing platform capabilities rather than a universal preference for one model over the other.

Nuclear power operator plans new design unit to support private players

Most Google Trends Are Already Over by the Time You See Them

A developer analysis of Google's Trending Now feed reveals that the majority of trending topics have already stopped trending by the time most tools or pipelines process them. Using four data fields — started_at, ended_at, is_active, and duration_minutes — it is possible to track the full lifecycle of a trend rather than just its rank. In a 24-hour snapshot of the US feed, 298 of 421 trending terms had already ended, with a median lifetime of roughly two hours. A 48-hour UK feed pull showed 530 of 615 terms had expired by the time of the pull, meaning pipelines that check once daily are reacting to largely stale data. The author recommends filtering on active status and sorting by duration as simple, low-cost improvements for any trend-driven workflow.

Nigeria's Kidnapping Crisis Spreads South, Leaving Communities in Fear

Nigeria is experiencing a growing kidnapping-for-ransom crisis that is expanding beyond its traditional hotspots into southern regions. Entire communities are living under the constant threat of abduction, with residents describing extreme danger in attempting to flee attackers. The crisis has intensified fears among civilians, who report that any attempt to escape during an attack can prove fatal. The spread of this criminal activity southward marks a significant and alarming shift in the country's security landscape.

SchemaCrawler adds SchemaSpy-compatible shim for gradual database doc migration

The latest SchemaCrawler release includes a compatibility shim that accepts standard SchemaSpy command-line arguments, allowing teams to switch tools without rewriting existing scripts. The shim, named 'schemaspy', is bundled in the SchemaCrawler Docker image and other installers, and works with multiple database types. Instead of SchemaSpy's HTML website output, it invokes SchemaCrawler Scribe to generate documentation in Google Open Knowledge Format, producing Markdown files with YAML frontmatter that can be stored and reviewed in Git. The Markdown-based output is easier to diff, search, and integrate with static-site generators or AI tools compared to browser-oriented HTML reports. Teams can use the shim as a transitional step before moving fully to SchemaCrawler's native command line and accessing its additional features like linting and row counts.

Why FutureBuilder and StreamBuilder Are Considered Flutter Anti-Patterns

Flutter's FutureBuilder and StreamBuilder widgets are widely used by developers to handle asynchronous data directly inside the UI layer, but experienced engineers argue this approach creates serious architectural problems at scale. Placing async logic inside a widget's build() method violates the core Flutter principle that UI should be a pure, synchronous projection of state, leading to unintended side effects. Common symptoms include duplicate network requests triggered by keyboard events, flickering sibling widgets, and full-screen reloads caused by simple user interactions like typing. The pattern also makes automated testing significantly harder, requiring complex workarounds with fake async timers and repeated pump calls. Experts recommend pushing asynchronous boundaries away from the presentation layer entirely, using state management solutions such as AsyncSignal, CubitSignal, or BlocSignal to keep widgets clean and testable.

Iceland Votes Saturday on Whether to Begin EU Membership Talks

Iceland is holding a closely contested referendum on Saturday to decide whether to pursue European Union membership negotiations. The vote comes amid rising economic pressures and shifting geopolitical tensions in the region. The outcome is expected to be extremely tight, with no clear frontrunner ahead of polling. A key sticking point for many Icelanders is control over their fishing waters, a vital part of the national economy. The result could mark a significant turning point in Iceland's long-standing relationship with the European Union.

Analysis of 1,033 dev launch posts finds 9 in 10 receive almost no engagement

A review of 1,033 posts on DEV.to found that roughly 90% of recent developer launch posts received three or fewer reactions and zero comments, a figure that held consistent across two separate observations. The analysis covered posts published within a ten-day window across eleven topic tags. Contrary to the author's initial expectation, low engagement did not correlate with weaker projects — several technically impressive tools were sitting at zero reactions. The investigation identified a recurring pattern: developers consistently buried their most compelling and distinctive feature deep within their posts, often as a minor bullet point. The core finding is that most launch posts open with generic descriptions while the one genuinely differentiated detail — the real reason someone would choose that tool — goes largely unnoticed near the end.

Rahul Gandhi's Remarks Reignite Debate on Gender Double Standards in Indian Politics

A fresh controversy has erupted in Indian politics after Rahul Gandhi made comments that have renewed questions about gender disparity in political discourse. The debate centres on whether women using strong or abusive language in politics is judged more harshly than when men do the same. The remarks were made in the context of insults directed at Prime Minister Narendra Modi. The episode has prompted widespread discussion about unequal standards applied to male and female politicians in India.

SEC Revives Crypto Custody Rule That Stalled Under Previous Administration

The U.S. Securities and Exchange Commission is working to revive a crypto custody rule that failed to pass under the previous administration. The original 2023 proposal sought to limit where investment advisers could hold clients' crypto assets. The new regulatory effort is still in early stages and details of the updated approach have not been made public. It remains unclear how the revised rule will differ from the earlier, narrowly scoped attempt.

India and Japan agree to review 15-year-old bilateral trade agreement

India has been pushing for a review of its trade agreement with Japan for several years, but progress was stalled due to resistance from the Japanese side. A significant step forward was made last month when the prime ministers of both nations agreed to revisit and modernize the pact. The two leaders acknowledged the need to make the agreement more relevant to current trade realities. Bilateral trade between India and Japan currently stands at $27.5 billion, with Indian exports accounting for approximately $6 billion of that figure.

Creatium Coach

Discussion | Link

Developer Builds Verifiable AI Agent Succession System Using Google Cloud Tools

A developer created Continuum, a system designed to ensure AI agents can hand off unfinished obligations to successor agents without losing verified context. The project was built for the All Things Agentic Hackathon and uses Google ADK, Gemini, and Cloud Tasks as core infrastructure. A synthetic €250,000 supplier-onboarding scenario was used as a reference case to test whether critical deadlines and obligations survive agent failure. The system produces five content-addressed control artifacts and uses a three-valued verification result — VERIFIED, FAILED, or INCONCLUSIVE — to distinguish genuine continuity from mere task replay. A secondary component called the Antibody Foundry generates multimodal outputs via Veo and Lyria only after a verifier confirms a VERIFIED result, enforcing a strict causal chain throughout the pipeline.

New Caledonia Launches HackAVP Hackathon to Improve Public Sector Job Access

HackAVP is a seven-week hackathon co-organised by OPT-NC, Station N, and the OPEN NC Data & AI commission, aimed at making public sector job listings in New Caledonia more accessible and user-friendly. The event kicks off on Wednesday, 9 September at 5:00 PM at Station N, with no prior registration required and teams formed on the day. The hackathon builds on a newly structured open data infrastructure in which OPT-NC's job vacancy notices are now normalised to schema.org/JobPosting format, exposed via an API, and accessible through an MCP server. Participants — including developers, students, designers, and HR professionals — are invited to build services that improve recruitment readability or find new uses for the public dataset. Finalists will present their projects live at TECH'N CO on 4 November, with all teams required to publish a documented article on DEV Community.

Meta's $18B Child Safety Settlement Raises Age Verification Privacy Concerns

Meta has reached an $18 billion settlement related to child safety, marking a significant legal milestone for the tech giant. A central condition of the deal involves the implementation of age verification technology on its platforms. However, experts and advocates have raised concerns that current age verification tools are unreliable and technically immature. The settlement has reignited a broader debate about the privacy risks that such verification systems pose to users. Critics worry that collecting age-related data could expose both minors and adults to new forms of privacy vulnerabilities.

Study Models Worst-Case Flood Risks From Himalayan Glacial Lakes

A 2022 scientific study published in Natural Hazards and Earth System Sciences examined worst-case glacial lake outburst flood scenarios in a transboundary Himalayan river basin. Researchers modeled the potential scale and impact of floods that could result from the sudden release of water stored in glacial lakes. Such events pose serious risks to downstream communities across national borders in the Himalayan region. The study contributes to growing scientific efforts to assess climate-related hazards in high-mountain environments shared by multiple countries.