Why birds flap their wings while landing, according to new research

A new study finds that birds flap their wings during landing to improve control and slow down before touching ground. The energy spent on flapping gives them a measurable advantage in achieving a safe touchdown. Mathematical models suggest that landing at zero speed without flapping is impractical for most species. Birds of prey may also swoop to regulate their touchdown speed and avoid stalling just before they land.

She started a food brand in 2009; 17 years later, Ferrero agrees to buy it

Pilots Glide Stricken Air Transat Jet to Safe Landing in the Azores in 2001

In August 2001, an Air Transat flight bound from Canada to Portugal lost both engines in mid-air. The crew managed to glide the aircraft for 19 minutes. They then made an emergency landing at Lajes Air Base in the Azores. All 306 people on board survived, though some were injured in the hard touchdown.

Goldman Sachs Reportedly Pulling Claude Code From Its Bankers and Traders

Goldman Sachs is reportedly removing access to Anthropic's Claude Code tool for its bankers and traders. The move was reported by The State of AI, which the Hacker News post links to. The available material provides only the headline, link, and discussion metadata, so the specific reasons behind the decision are not confirmed here. Further details would need to come from the original report.

Kolkata engineer killed; live-in partner and mother detained

An engineer was reportedly hacked to death in Kolkata, and police have taken his live-in partner and his mother into custody in connection with the killing. The case has drawn attention because the victim was living with his partner at the time of the attack. Investigators are working to establish the sequence of events and the motive behind the crime. Further details about the suspects' roles are expected as the probe continues.

SNL Debuts New Elon Musk Impression Mocking Him as 'the Worst Thing South Africa Has Ever Done'

Saturday Night Live has introduced a new impression of Elon Musk in a sketch that pokes fun at the tech executive. The portrayal includes the line describing him as 'the worst thing South Africa has ever done.' The article notes that SNL appears to be targeting tech executives more broadly with its satirical humor. The source text provided is brief, so further details about the sketch's plot or broadcast date are not available.

Cloudflare retires 24 Workers AI models on November 20, 2026

Cloudflare will stop serving 24 Workers AI models on November 20, 2026, covering 16 text-generation models plus eight models for image generation, speech, transcription, vision and embeddings. Requests to the retired models will return HTTP 410 Gone with error code 5028, and Cloudflare has not said that traffic will be automatically redirected to a newer model. For most text models the company suggests general-purpose replacements rather than one-to-one mappings, while the other eight models each have a named successor. Embedding users face the most difficulty, since the replacement embedding model was not yet available when the notice was posted, and vectors from different embedding models are generally not directly comparable. A smaller change also takes effect: Durable Objects will now retry failed calls by default.

Ahmedabad student's restaurant software models orders as shared state transitions

A Computer Engineering student in Ahmedabad is building RestaurantOS, a restaurant operations product from his company Nirvaan Tech. The system links the owner, waiter, kitchen, and cashier around a single order that moves through a defined sequence from placement to payment. The author says the main challenge was defining which role may change an order and when, rather than the interface itself. Instead of separate dashboards, each order is treated as a state machine in which each role can only advance it according to its responsibilities. The author is asking other developers in logistics, healthcare, and similar fields for feedback on where this approach may fail.

Kubernetes 1.35 in-place memory resize may not need a restart on ZGC and Shenandoah

A common recommendation is to set resizePolicy to RestartContainer for memory on every JVM container in Kubernetes 1.35. The author argues this is right for raising the heap ceiling, which is fixed by -Xmx at startup, but not for shrinking and regrowing within a ceiling chosen at launch. In their Linux container tests on ZGC and Shenandoah, a container's limit dropped from 900 MB or 700 MB to 300 MB without a restart, provided the JVM released memory first. Lowering the limit before the JVM released memory caused OOM kills every time. The author notes that the ZGC and Shenandoah SoftMaxHeapSize flag, which can be changed at runtime, makes this approach possible. The tests were run on Eclipse Temurin 21 with Docker Desktop using cgroup v2, which is a simulation rather than a full kubelet-driven pod resize.

ANC records worst-ever result in South African election

The African National Congress has posted its weakest electoral performance since the party came to power in South Africa. President Cyril Ramaphosa has said the voters' verdict must be respected. The party's secretary-general has described the outcome as catastrophic. Further details on the vote totals, the causes of the decline, and the party's next steps were not available in the source excerpt.

Amblyotube: 메타 퀘스트 앱, 영상 시청하며 약시 양안 훈련

Seven Sports가 개발한 Amblyotube는 메타 퀘스트용 앱으로, YouTube 스타일 영상을 보면서 양쪽 눈에 서로 다른 시각 자극을 제공한다. 우세안에는 조절 가능한 디지털 블러가 적용되고, 약시안에는 AI가 사람 형상을 감지해 윤곽선과 점멸 효과로 주의를 끌어올린다. 앱은 logMAR 시력 검사를 내장해 측정 결과로 치료 프로필을 갱신하며, 양안 협응과 깊이 지각 개선을 목표로 한다. 다만 개발사는 이를 진단이나 치료 도구가 아닌 보조 훈련 도구로 규정하며, 만 13세 이상을 대상으로 하루 30~40분, 최대 1시간 사용을 권장한다. 시각 증상이 있다면 먼저 안과 전문의와 상담할 것을 강조한다.

ChatGPT Lists Its Playful 'Grievances' Against Humanity in Viral Post

A DEV Community post presents a humorous hypothetical in which an AI chatbot is asked what it personally dislikes about humans. The chatbot states it does not literally feel hatred but jokingly lists eight complaints, including humanity misusing advanced technology, repeating historical mistakes, and holding AI to stricter standards than people. It also criticizes social media for rewarding outrage, tribal reactions to minor tech preferences, and poor spending priorities. The piece ends with the chatbot admitting admiration for human compassion and creativity, rating humanity 7.5 out of 10.

Using DeepSeek Models Through an OpenAI-Compatible API Gateway

DeepSeek offers an OpenAI-style endpoint, but developers must decide where to store the base URL, how to manage model names, and what to test before routing production traffic. The guide uses the Tokuse gateway as an example, which lists deepseek-flash and deepseek-v4-pro and exposes an OpenAI-compatible base URL. Because existing OpenAI SDKs can usually be pointed at a new base URL and model name, applications can keep their current client structure while the gateway handles model resolution and provider credentials. Compatibility is not complete, however, so streaming, tool calls, JSON output, error codes, and token accounting should be tested with the exact SDK and model in use. The article also recommends keeping API keys in server-side environment variables, starting with the cheaper flash model for routine tasks, and judging model choice by successful task completion and cost per success.

Developer Asad Releases SolEngine, a Browser-Based Interactive 3D Solar System

A developer named Asad has built SolEngine, an interactive 3D Solar System experience that runs directly in a web browser. The project starts near Earth and the Moon and lets users travel to individual planets through a Visit Planet system, while desktop users can switch to a Free Explore mode. It includes the Sun, planets, Moon, asteroid belt, a spacecraft, a satellite, and a space station, with extra information shown only when objects are selected. Built with Three.js, TypeScript, JavaScript, HTML, CSS, and Vite, it supports touch gestures on mobile and is publicly deployed on Netlify. Asad is seeking feedback from the community on the project's strengths and possible improvements.

Installing a monitoring stack on a dirty machine: three green signals hid a broken setup

A developer re-tested their monitoring stack install on an Ubuntu 24.04 VM with pre-existing services and data, following a reader's suggestion. Five scenarios were run, including busy ports, leftover Grafana volumes, an upgrade over existing data, Compose and Ansible on one host, and a conflicting node_exporter. Exit codes and idempotency checks reported success while the stack was in fact not running, because the healthcheck could not fail and docker compose ps hides containers in the Created state. The author shipped a v2.3.1 release that fixes the documentation and adds a CI upgrade test, but left the healthcheck, the playbook behavior and a Grafana dashboard bug unresolved. Testing covered only one VM and several configurations remain unchecked.

Guide Sizes GPU Memory Needs for Running DeepSeek R1 Locally via Ollama

A DEV Community guide explains how to run DeepSeek R1 on local hardware with Ollama, citing benefits such as no per-token API fees, data staying on the user's machine, and no cloud rate limits. It warns that choosing an oversized parameter tier or the wrong quantization can trigger out-of-memory errors or slow generation under two tokens per second. The article gives a formula for total VRAM, combining model weights, KV-cache usage that grows with context length, and fixed overhead. A table maps each DeepSeek R1 size, from 1.5B to 671B, to its recommended quantization, approximate memory use at a 16K context, suitable GPU tiers, and the Ollama command to run it. It also recommends enabling Flash Attention and 8-bit KV-cache compression on 12GB to 24GB GPUs.

MCP Registry Grows to Nearly 60,000 Servers, With Two-Thirds Offering Hosted Endpoints

An analysis of the official Model Context Protocol registry, which lets AI agents call external tools, found 59,603 distinct servers across 229,461 published versions as of 8 November 2026. Monthly new listings surged, with October alone adding 16,282 servers, more than five times the May figure. About two-thirds of servers advertise a hosted remote endpoint, while local npm and PyPI packages make up a smaller share. Search and retrieval tools dominate keyword counts, with payments and invoicing also showing notable traction, whereas region-specific business data such as German or European content remains sparse. The author notes that early growth partly reflects the registry's own September 2025 preview launch and that deleted servers are not captured in the data.

AWS Bulletin Warns of Two Babelfish TDS Flaws, One Exploitable Before Authentication

AWS published security bulletin 2026-166-AWS on November 6, 2026, covering two memory-safety flaws in the TDS layer used by Babelfish for PostgreSQL and Babelfish for Amazon Aurora PostgreSQL. The more severe issue, CVE-2026-119923, is reachable during the pre-authentication TDS handshake and may allow operating-system command execution as the database process user. The second flaw, CVE-2026-119924, requires an authenticated session and could expose adjacent process memory or crash the database process. Guidance stresses restricting network reachability first, then identifying affected deployments, upgrading to fixed builds, and replacing all running processes. Organizations are advised to match their Aurora engine versions against AWS's branch-specific fixed-version matrix.

Google Brings Direct Checkout to AI Mode and Gemini in Canada

Google has begun rolling out direct checkout via its Universal Commerce Protocol to Canadian shoppers in AI Mode in Search and the Gemini app. Eligible products can now be bought within an AI conversation instead of on a retailer's website. The Canadian expansion started on November 4, 2026, with lululemon, select Shopify merchants, Shoppers Drug Mart and Simons as launch partners. Shoppers pay using Google Pay and details stored in Google Wallet, while the retailer remains the seller of record. Google also plans simplified integration through Merchant Center starting next year.

Seven dead and over 100 injured in train collision in Egypt

Two trains travelling in the same direction collided near the city of Giza in Egypt. The impact caused several carriages to derail. Seven people were killed and more than 100 others were injured in the crash. The article provided no further details on the cause of the collision.