SShortSingh.
0
TechnologyThe Verge ·

Nvidia Launches Probe Into CEO Jensen Huang's Personal Relationship

Nvidia's board has opened an investigation into chief executive Jensen Huang's relationship with Alice Huang, an AI executive described as an agent of the Chinese government. The probe was announced on Sunday. Huang also sent a memo to employees acknowledging the affair and apologizing to his family and colleagues. In the memo he took full responsibility and said he had been unfaithful to his wife, with the actions hurting her and their children. The source article provides the full memo text.

0
ProgrammingDEV Community ·

Daily Payment Reconciliation Scheduling Relies on Idempotent Run Keys and Webhook Retries

A DEV Community write-up argues that scheduler choice should be based on the delivery guarantees it can actually prove, not on the convenience of its interface. For a healthtech payment reconciliation job, the author recommends at-least-once execution paired with an idempotent report key. The webhook is treated only as a trigger, while an immutable run record serves as the authoritative source of truth. The article proposes scheduling a run at 02:00 UTC, with separate boundaries tracked for scheduler delivery, run claiming, report publishing, and consumer acknowledgement. It also explains that an idempotency key such as the business date, rather than an "exactly once" label, lets repeated webhook deliveries map to a single logical run.

0
ProgrammingDEV Community ·

Recurring HTTP Triggers Can Drive Bounded Data Cleanup in B2B Systems

The article proposes using an external recurring scheduler to send a small, authenticated HTTP request to a cleanup endpoint instead of running an internal task executor. The endpoint claims a bounded time window and delegates the actual processing to a separate worker pool, keeping the trigger stateless. Shorter intervals reduce cleanup latency but raise request pressure, so the endpoint should cap concurrency, honor Retry-After headers, and make retries safe. The author suggests a starting policy of a five-minute tick with a one-minute window and a two-minute safety delay for clock skew and late writes. Responses such as 202 Accepted or 429 Too Many Requests are recommended to signal accurate status to callers.

0
ProgrammingDEV Community ·

Study Compares 500- and 1500-Token Chunks for Marketplace Handbook Retrieval

A DEV Community article examines how chunk size affects retrieval quality when answering questions from a marketplace product handbook. It argues that 500-token chunks can isolate a relevant clause but drop its heading or exceptions, while 1500-token chunks keep more context but may bury the matching rule among unrelated text. The author recommends splitting along heading-bounded sections first and comparing fixed sizes against that baseline using the same questions and citation requirements. The piece also stresses tracking document revision and chunk identity so citations resolve to the revision used, and proposes a 30-question evaluation fixture covering direct lookups, cross-section questions, and unanswerable cases. The article treats the problem as one of grounding answers in verifiable evidence rather than simply tuning chunk size.

0
ProgrammingDEV Community ·

Debugging DNS Resolution Failures When the Server Itself Is Reachable

A server can answer pings and still be unreachable by name because DNS resolution is failing, which is often the true cause of the problem. The author outlines a practical troubleshooting sequence, beginning with dig and nslookup to determine whether the resolver or the remote host is at fault. Common culprits include an overwritten /etc/resolv.conf pointing to a dead nameserver, an unreliable internal resolver, systemd-resolved stub confusion, stale or poisoned caches, and firewall rules that block UDP port 53. To prevent recurrence, the author recommends pinning a known-good resolver configuration in systemd-resolved and monitoring DNS query latency alongside ping checks. The full postmortem with packet captures is published on the author's ops notes site.

0
ProgrammingDEV Community ·

Automating Custom Golden VM Images with CDI and OpenShift Virtualization

Traditional VM provisioning from ISO files is too slow for cloud-native platforms, so OpenShift Virtualization uses pre-built golden images that can be deployed in seconds. The described pipeline has three stages: packaging a qcow2 disk into a lightweight container image built from scratch, storing it in the OpenShift Internal Registry, and having CDI periodically import updates into a DataSource for cloning. Setup involves setting restrictive file permissions on the disk, enabling the registry's default route, and pushing the image with Podman. CDI's importer is configured with registry credentials and the service CA certificate to avoid x509 errors, and a DataImportCron resource with an immediate-binding annotation schedules recurring imports every 15 minutes.

0
ProgrammingDEV Community ·

Codex CLI가 osFoundry 세션에 작업을 넘기는 연결 설정 방법

Codex CLI는 자신이 직접 수행하지 못하는 작업을 osFoundry의 에이전트, 팀, 루프 세션으로 넘길 수 있게 되었다. 넘기는 항목은 목표, 지금까지의 파악 내용 요약, 원하는 결과물이며 세션 오너에게 알림이 가고 오너가 승인하거나 거절한다. 승인된 작업이 끝나면 Codex는 같은 터미널에서 결과를 읽고 이어서 작업한다. 설정은 osFoundry에서 Connected 탭으로 Codex를 등록하고, config.toml에 필요한 권한 범위와 툴만 지정한 뒤 로그인하는 순서로 진행된다. 세션에 대한 역할은 Viewer, Contributor, Runner로 나뉘어 읽기, 노트와 핸드오프 제안, run 요청 권한을 각각 정한다.

0
ProgrammingDEV Community ·

Meta to End Quest 2 and Quest Pro Support; Amblyotube Vision Training App Explained

Meta has officially announced the end of support for the Quest 2 and Quest Pro headsets, with the final OS update arriving early next year before development shifts to next-generation hardware. Seven Sports, the developer behind the app, has released Amblyotube, a Meta Quest application that delivers different visual input to each eye while users watch YouTube-style content. The app is designed as a supplementary tool for binocular coordination and depth perception, aimed at people with amblyopia (lazy eye). Its developers say it uses adjustable intensity presets, AI-based outlining and a fixed flicker rhythm, with the aim of keeping daily training engaging. The company stresses that Amblyotube is not a diagnostic, therapeutic or curative device, recommends 30 to 40 minutes of use per day, and sets a minimum age of 13.

0
ProgrammingDEV Community ·

Six-stage evidence pipeline aims to make enterprise agentic RAG answers more reliable

A solutions architect's assistant produced a polished brief citing an outdated RFP revision and a revenue figure taken from a slide deck rather than the finance system, showing that a grounded answer can still use the wrong evidence. The article argues teams should first examine retrieval architecture, asking what evidence the model could see, where it searched, and how recently it was updated, before switching models. It defines agentic RAG as a system where the model chooses tools and retries when evidence is weak, while the application keeps control of security, access, validation, and stopping conditions. The proposed design is a six-stage pipeline (prepare, plan, route, select, assemble, answer) with a continuous data-preparation stage and per-question stages that loop back within set limits. The piece also distinguishes indexed retrieval, which is fast but may be stale, from live retrieval, which is needed for current figures, and stresses tracking version, ownership, and permissions as metadata.

0
ProgrammingDEV Community ·

Auditoría en código de una teoría física de tiempo cuantizado de Ricardo Cergneux

Un prototipo de código abierto en Python llamado cergneux-audit-demo convierte las afirmaciones de dos documentos del físico Ricardo J. Cergneux en pruebas automatizadas. Cada test se vincula a una página concreta del texto y registra un veredicto sobre si la afirmación se sostiene. Según el autor, algunas afirmaciones resultan correctas, otras son identidades disfrazadas y otras no superan la comparación con datos. Entre los hallazgos figuran un signo incorrecto en una ecuación del hamiltoniano Josephson y una discrepancia de unos 49 órdenes de magnitud en una estimación temporal. El autor presenta la herramienta como una demo v0.x abierta a correcciones, no como un veredicto definitivo sobre toda la obra del investigador.

0
ProgrammingDEV Community ·

FLUX 3 Klein Runs on 8 GB VRAM in ComfyUI, but the Upsampler Adds 9.8 GB

Black Forest Labs' FLUX 3 Image Klein, released October 15, 2026, can run locally in ComfyUI on an 8 GB card such as the RTX 3070, though the required prompt upsampler is itself a 9.8 GB file. The full int8 download totals about 20.8 GB of weights, which ComfyUI loads in stages and moves to system RAM when not in use. The model comes in a distilled checkpoint for generation and editing and an undistilled Base checkpoint for fine-tuning, each requiring different settings. BFL recommends always using the upsampler, since short prompts produce generic output, and the author estimates 32 GB of system RAM as a practical minimum. The article also notes that the company's reported Elo gains are internal evaluations not yet independently verified.

0
ProgrammingDEV Community ·

Choosing Storage for Browser Receipt Uploads: Retention, CORS, and Deletion Trade-offs

The article argues that Cloudflare R2 is the strongest fit among R2, S3, and Wasabi for browser-uploaded receipts with meaningful download traffic, citing low storage cost, free egress, and CORS support via presigned URLs. Amazon S3 is presented as better for global coverage and fine-grained controls, while Wasabi is flagged as risky where frequent native egress is needed. It stresses that cheap uploads are not a retention policy, so applications still need deletion schedules, audit records, and clear ownership of removals. The recommended pattern has the backend authenticate users and issue short-lived presigned upload URLs, with metadata stored in the application database rather than the storage provider. Finally, it notes that CORS must be verified at the bucket layer and proposes a five-case test to compare providers before committing.

0
TechnologyThe Verge ·

Soul Glo's Diaspora Problems stands out as an unusual hardcore record

Soul Glo has a new album, NA I'M GON GRAB ALLAT, scheduled for release in January, and two singles have already been issued. While the writer awaits it, they have been revisiting the band's 2022 album Diaspora Problems, which they name as their favorite record of that year. The article highlights the opening track "Gold Chain Punk (whogonbeatmyass?)", describing a hardcore riff and frontman Pierce Jordan's shouted, intense delivery. It also notes a half-time stomp section where Jordan's vocals grow more frenzied, and says the energy stays high as the song shifts between sections.

0
ProgrammingHacker News ·

KetchUp: New Open-Source SketchUp Alternative Built with Rust and WebGPU

A new open-source 3D modeling application called KetchUp has been introduced as an alternative to SketchUp. The project is built using the Rust programming language and the WebGPU graphics API. Its project page is hosted on GitHub Pages and was shared on Hacker News, where it gathered 14 points and 2 comments. The source article provides limited detail beyond the project link and community discussion, so specific features and release status are not described here.

0
IndiaTimes of India ·

16-year-old Vedant Thakker builds free app to flag wildfire risk for Californians

Vedant Thakker, a 16-year-old, has developed a free application called Red Flag Check. The tool assesses wildfire danger using data from the National Weather Service. It sorts residential locations under red flag warnings into four distinct risk levels. The app is intended to give local residents clearer information about their exposure. It is designed with older adults in mind, who may be more vulnerable during wildfires.

0
TechnologyNYT Technology ·

Nvidia Investigates Its CEO Over Admitted Relationship With Alleged Chinese Spy Recruit

Jensen Huang, chief executive of the A.I. chip maker Nvidia, has acknowledged a romantic relationship with a woman. U.S. intelligence officials say she was recruited by China's top intelligence agency. The admission has prompted Nvidia to open an investigation into its chief executive. The article's source text provides limited detail beyond these core facts, so the full scope of the probe and its findings remain unclear. The matter highlights ongoing concerns about foreign espionage targeting leading U.S. technology companies.

0
IndiaTimes of India ·

Former Manhattan Firehouse Sells as Luxury Townhouse for $19 Million

A historic fire station in Manhattan has been converted into a luxury townhouse and sold for $19 million. The building served its neighborhood as a firehouse for more than a century before it was decommissioned. Extensive renovations were carried out on the property, which features an underground jazz club among its distinctive attributes. The townhouse keeps reminders of its industrial origins while adding contemporary living areas.

0
ProgrammingHacker News ·

Nvidia CEO Jensen Huang addresses staff in letter about affair

Nvidia CEO Jensen Huang reportedly sent a letter to employees concerning an affair. The report was published by Axios and gained traction on Hacker News, where it drew 79 points and 27 comments. The article text provided is limited to links and engagement metrics, so the letter's full contents and specific details cannot be confirmed from this source. Further reporting would be needed to establish the exact nature of the message and its circumstances.

← NewerPage 433 of 7347Older →