Candidats 218
Bruts 512
Sources 26/26
Collecté le 2026-10-20 10:25 UTC
Fenêtre depuis 2026-10-13 00:00 UTC
[1] Cloudflare Blog — efficience — économie du calcul (11)
- 2026-10-19Cutting our inference bill by 68% by being obsessive about optimization
- 2026-10-19Introducing Clairic, a new System One class of extraction models
- 2026-10-19Beyond the patch: Protecting your NetScaler gateway against future vulnerabilities
- 2026-10-16Continuous accountability for Verified bots
- 2026-10-15A fully programmable secure web gateway
- 2026-10-15Deploy with Void
- 2026-10-15Cloudflare and OpenAI: Making AI discovery better for everyone
- 2026-10-14Yandex outage: Data shows that instead of going dark, malicious bots went on the move
- 2026-10-14Inside Hyperdrive: How we solved a rare Hyperdrive MySQL caching bug
- 2026-10-13Under the hood, off the hot path: How D1 scales to millions of databases
- 2026-10-13Rewinding the exponential: How we keep hardware usage down
[1] GitHub Blog — efficience — outils (6)
- 2026-10-19How to continue a GitHub Copilot app session in VS Code
- 2026-10-15The next developer interface is ambient intelligence
- 2026-10-15Staying Evergreen: How we automated secure access and repo hygiene in a shared org
- 2026-10-14GitHub availability report: September 2026
- 2026-10-14From 2,500 requests to 24: Rebuilding star history without the stargazers
- 2026-10-13GitHub’s model for secret detection: Finding more secrets with less noise
[1] GitHub Changelog — efficience — outils (10)
- 2026-10-19Copilot model policy API for enterprise, orgs, and teams
- 2026-10-19GitHub Copilot models deprecated on October 19, 2026
- 2026-10-19Some Git over HTTPS may temporarily require authentication
- 2026-10-19CodeQL changes for macOS 27, Xcode 27, and Swift 6.4
- 2026-10-16Open source license compliance generally available
- 2026-10-16Xcode 27 runner image generally available
- 2026-10-16Multi-repository support and other improvements for GitHub in Slack and Teams
- 2026-10-15Download GitHub Code Quality license usage as a CSV
- 2026-10-15Upcoming deprecation of selected GitHub Copilot models in mid-November
- 2026-10-15AI Scan for pull requests generally available
[1] InfoQ — efficience — pratique d'ingénierie (15)
- 2026-10-19Java News Roundup: JEPs for JDK 28, Azul Payara, Gatherers4j, Google ADK, Yupiik, Apache Camel CVEs
- 2026-10-19Pinterest Uses Event Time Watermarks to Finalize Petabyte Scale Data Partitions
- 2026-10-19JEP 543 Finalizes Structured Concurrency in JDK 28, Ending Seven Preview Rounds
- 2026-10-19Article: From Microservices to Agents: the Next Evolution of Distributed Systems
- 2026-10-19Podcast: The Future of Web Apps: Will AI Agents Generate UIs on the Fly?
- 2026-10-19EmbeddingGemma 2 Expands beyond Text to Support Code, Images, Video, and Audio
- 2026-10-19Buildkite Adds Scoped Secrets and New Agentic CI Tools
- 2026-10-19.NET 11 Release Candidate 2: Go-Live License and Updates across Libraries, F#, MAUI, and WinForms
- 2026-10-19NestJS 12: ESM Packages, Standard Schema Validation and Native Observability
- 2026-10-18Google Adopts Pyrefly for Faster, More Accurate Python Type Checking
- 2026-10-18AWS Introduces Preview of Well-Architected Agent
- 2026-10-18AWS Will Run Its European Sovereign Cloud Without the Global Backbone for Several Hours
- 2026-10-17Reverse Engineer Anything Project Brings AI Agents to Binary Analysis
- 2026-10-17Amazon Acquires DuckLabs and Integrates DuckDB in Amazon Aurora PostgreSQL
- 2026-10-16Google EnvHarness Enables Self-Evolving Agents to Learn in Dynamics Environments
[1] Meta Engineering — efficience — ingénierie à l'échelle (6)
- 2026-10-19GenRecs Tokenization: Transforming Ads Recommendation at Meta with Stable Model Vocabulary
- 2026-10-19The Thin Orchestrator: How We Made Our AI Agents More Reliable by Giving the Router Less to Do
- 2026-10-19Inside Prometheus: Meta’s Journey to a Gigawatt-Scale Supercluster
- 2026-10-15How Meta Built Forum With AI-Native Engineering
- 2026-10-152026 World Cup: Breaking Records, Spiky Charts and Reliability at Scale
- 2026-10-13Meta’s Generative Ads Model (GEM): Scaling the Central Model and Compute Engine for Frontier Ads AI
[1] Netflix Tech Blog — efficience — ingénierie à l'échelle (10)
- 2026-10-16Ontology-Driven Observability: Building an End-to-End Knowledge Graph at Netflix Scale
- 2026-10-14Integrating Netflix’s Foundation Model into Personalization applications
- 2026-10-14DataJunction as Netflix’s answer to the missing piece of the modern data stack
- 2026-10-14Measuring the Impact of Personalized Recommendations
- 2026-10-14TimeSeries Tiered Storage Journey: Kafka/Flink Streams to Native Cassandra Cold Reads
- 2026-10-14Behind the Scenes: Evolving Netflix’s Ads Event Pipeline for Live — Part II
- 2026-10-14Netflix Conductor: The Next Chapter
- 2026-10-14Running Apache Spark experiments in my sleep (and on a plane)
- 2026-10-14The Lifecycle of LLM-as-a-Judge: Building, Aligning, and Monitoring at scale
- 2026-10-14Netflix Stratum Media Processing: Automated Container Right-Sizing
[2] Google DeepMind — modèles & coût (5)
- 2026-10-15Accelerating the search for antimicrobial peptides with Google Antigravity
- 2026-10-14Google DeepMind and Singapore Government collaborate to build fine-tuned AI model for scam detection
- 2026-10-13Developing Gemini 4 Argon responsibly and safely
- 2026-10-13Gemini 4 Argon for Robotics
- 2026-10-13Gemini 4 Argon: Built to solve your hardest problems
[2] Hugging Face Blog — modèles & coût (9)
- 2026-10-19TorchNeuron: running HF models on AWS Trainium is easier than ever
- 2026-10-16Scaling Indic Pretraining: 2.8 Trillion Tokens and an Open Data Generation Pipeline
- 2026-10-16Welcome Beam: a 501B open model you can own, customize, and scale
- 2026-10-15Granite Switch 4.1: Chat, Decisions, and Speech from One Endpoint
- 2026-10-15Apertus 1.5 Release Notes
- 2026-10-15Welcome Muse Spark Open: open-weight frontier multimodal mixture-of-experts model from Meta
- 2026-10-15Welcome FLUX.3 Image Klein in Diffusers
- 2026-10-14HiRO-ACEv2: Simulating extreme weather at 3 km resolution for a fraction of the cost
- 2026-10-13Run fast, private agents on-device with LFM2.5-24B-A2B
[2] NVIDIA Blog — modèles & coût (9)
- 2026-10-20Spain-Based Startups Advance Regional Innovation With Barcelona Supercomputing Center
- 2026-10-19Germany Wants to Lead the World in Industrial AI: How NVIDIA Will Help
- 2026-10-17RTX Spark Ignites Global Celebration as New Windows PCs Arrive
- 2026-10-15Open From Data Center to Desktop: NVIDIA Expands Nemotron Family With Highest Efficiency Super VL and Bolt Models
- 2026-10-15Dial Up the Cloud: NVIDIA DLSS 4.5 Controls Arrive on GeForce NOW
- 2026-10-15NVIDIA Expands Advanced Technology Access for Europe’s Scientific Community Through Open Learning Resources, Collaborations and Compute
- 2026-10-15The Nordics and NVIDIA Show What’s Possible When a Region Commits to Advanced Computing
- 2026-10-14NVIDIA Announces New Investments in the UK Startup Ecosystem
- 2026-10-13All Systems Go: Global PC App Ecosystem Ready for NVIDIA RTX Spark
[2] OpenAI News — modèles & coût (11)
- 2026-10-19Advancing coding in low‑resource programming languages
- 2026-10-16How Itaú is building generative AI for its entire customer base
- 2026-10-16Stack Overflow for Agents triples signups with ChatGPT plugin
- 2026-10-15Teaching students to think and learn through questions
- 2026-10-15Sullivan & Cromwell brings legal expertise into agreement review
- 2026-10-14Amp puts AI agents to work beyond code with OpenAI models
- 2026-10-14SentinelOne cuts malware analysis from weeks to hours
- 2026-10-14How Third Moment Research runs 10,000 agents for 36+ hours
- 2026-10-14How Block is building AI-powered security with OpenAI technology
- 2026-10-14The EU’s next chapter in AI is agentic
- 2026-10-13Higgsfield takes AI video from 1 minute to over 10
[3] CNCF Blog — infra & plateformes (3)
- 2026-10-19How AI workloads are reshaping Kubernetes autoscaling
- 2026-10-16Deploying, maintaining, and migrating production n8n on EKS
- 2026-10-15Extending Kubernetes policy enforcement to runtime security for AI workloads
[3] Kubernetes Blog — infra & plateformes (1)
[3] LWN.net — infra & plateformes (15)
- 2026-10-19Linux Plumbers Conference videos now available
- 2026-10-19[$] Development statistics for the 7.3 kernel
- 2026-10-19[$] Cross-crate forward declarations in Rust
- 2026-10-19Security updates for Monday
- 2026-10-18The 7.3 kernel has been released
- 2026-10-16Introducing GNOME OS Preview
- 2026-10-16[$] Safe self-referential data structures in kernel Rust
- 2026-10-16[$] Recent and future changes to the Linux NFS server
- 2026-10-16Results from the 2026 TAB election
- 2026-10-16Security updates for Friday
- 2026-10-15[$] Vendor hooks in the Android kernel
- 2026-10-15KDE releases documentary
- 2026-10-15Ubuntu 26.10 released
- 2026-10-15Security updates for Thursday
- 2026-10-15Seven stable kernels for Thursday with fix for critical vulnerability
[4] Martin Fowler — pratique & validation (4)
- 2026-10-15When Software Meets Hardware
- 2026-10-14Build your slide template
- 2026-10-13The Craft of Practitioner Voice
- 2026-10-13Are Harnesses Worth the Cost?
[4] Simon Willison — pratique & validation (15)
- 2026-10-19My obituary for programming
- 2026-10-18llm-tools-datasette 0.1.1
- 2026-10-18llm-echo 0.5
- 2026-10-17Inline rendering with document.currentScript
- 2026-10-17web-edit 0.1a2
- 2026-10-16Quoting Addy Osmani, Morgan Westlee Lunt, and Qing Wang
- 2026-10-15Chirp Pong
- 2026-10-15Anthropic tells investors quarterly revenue up nearly 50% to $16.8bn ahead of IPO
- 2026-10-15LaTeX renderer
- 2026-10-14Tree Appreciator
- 2026-10-14AI is mass decompiling games. That's both awesome and scary.
- 2026-10-13Redesigning source control at Anthropic
- 2026-10-13llm-gemini 0.35
- 2026-10-13Quoting Geoffrey Huntley
- 2026-10-13web-edit 0.1a1
[4] Stack Overflow Blog — pratique & validation (8)
- 2026-10-20Don't forget about the UI for AI
- 2026-10-19Production-grade LLMs and agents: a field guide
- 2026-10-16AI says you’re right, formal methods say you’re wrong
- 2026-10-15MCP standardized tools. Now we need the infrastructure around them.
- 2026-10-15Why your platform team shouldn't build an AI context layer (even though they probably could)
- 2026-10-14Your next job doesn’t have a title yet: Emerging role skills Q&A with Cory Hymel of Andela
- 2026-10-13Why Your Sandbox Never Quite Matches Production (And Why That's Fine)
- 2026-10-13Keep calm and call clone with Rust
[4] The New Stack — pratique & validation (26)
- 2026-10-19MiniMax just halved its token prices. Your AI agent bill may not follow.
- 2026-10-19One sentence changed Claude Code’s overreach rate from 0% to 17.1%
- 2026-10-19DeepSeek made its name with cheap AI models. Now it’s undercutting search API rivals, too.
- 2026-10-19The distroless security gap: Why bottom-up containers are the next evolution in OCI images
- 2026-10-18Gemini 4 Argon vs. Claude Opus 5.5: Half the price and about half as accurate
- 2026-10-18Liquid AI caught its model cheating on coding tests — then it recognized the sandbox
- 2026-10-18Scale isn’t one number: three ways a Kubernetes control plane runs out of room
- 2026-10-17Here are the engineering lessons we learned when live-streaming sports to 32 million people
- 2026-10-17AI has lowered the barrier to building. Here’s what that means for engineers.
- 2026-10-17His scanner wouldn’t work on Linux — so he let Claude study the Windows driver
- 2026-10-17AI agents vs. Kubernetes: where each belongs in the control loop
- 2026-10-16AI agents turn one request into about 20. Kubernetes is evolving to handle it.
- 2026-10-16Only 34% trust synthetic data’s referential integrity — and it’s holding adoption back
- 2026-10-16Employees are vibe-coding their own apps. Cloudflare open-sources Void to give IT a handle on them
- 2026-10-16Running local AI models on Microsoft’s Surface Laptop Ultra: What we found
- 2026-10-16Muse Wonder vs. Claude Fable 5.1: Meta’s new model came closer than I expected
- 2026-10-16“Developers are paying frontier prices for amnesia”: Uvi wants to stop the agent relearning brain drain
- 2026-10-15Hackers hijacked 1,900 GitHub accounts. Then they went after 30,000 repos.
- 2026-10-15Vercel adds always-on workers to run more of your backend
- 2026-10-15OpenAI promised developers six months’ notice. GPT-4o users got three.
- 2026-10-15Meta ships Muse Spark Open, its answer to China’s best open-weight models
- 2026-10-15Nvidia squeezed a 74B agent model into 24GB so it can run on your laptop
- 2026-10-15What it takes to become an AI company: Four lessons from the top
- 2026-10-15Anthropic’s been eyeing confidential AI. It just made its move with another acquihire.
- 2026-10-15“Everybody with a $20 account is now a researcher”: How Akrites is tackling the flood of AI bug reports
- 2026-10-15Anthropic built its own git replacement for AI-scale coding. But you can’t have it.
[5] CISA KEV — impact de masse — sécurité à impact collectif (7)
- 2026-10-16CVE-2026-102764 — BeyondTrust Remote Support: BeyondTrust Remote Support Authentication Bypass by Spoofing Vulnerability
- 2026-10-16CVE-2025-61624 — Fortinet Multiple Products: Fortinet Multiple Products Path Traversal Vulnerability
- 2026-10-15CVE-2026-21589 — Atlassian Multiple Products: Atlassian Multiple Products Files or Directories Accessible to External Parties Vulnerability
- 2026-10-14CVE-2026-76519 — Cisco Secure Email Gateway: Cisco Secure Email Gateway Relative Path Traversal Vulnerability
- 2026-10-14CVE-2026-0322 — Palo Alto Networks PAN-OS: Palo Alto Networks PAN-OS Code Injection Vulnerability
- 2026-10-13CVE-2026-104935 — Ivanti Connect Secure: Ivanti Connect Secure Authentication Bypass by Spoofing Vulnerability
- 2026-10-13CVE-2026-96951 — Microsoft Windows: Microsoft Windows Command Injection Vulnerability
[5] GitHub Advisories critiques — sécurité à impact collectif (10)
- 2026-10-19GHSA-f8c7-rwf9-fmgm [critical] Mercury: Unauthenticated Remote Code Execution in /api/terminals - access control bypass
- 2026-10-19GHSA-rv4m-xm3f-3ccq [critical] Mercury: Unauthenticated Remote Code Execution via /api/kernels
- 2026-10-16GHSA-w5f2-9q9c-ff4h [critical] LMCache: Remote code execution via unsafe deserialization in multiprocess mode transport
- 2026-10-16GHSA-h4mw-j7qp-8mwm [critical] Claude Agent SDK: Argument Injection via resume Option Allows Arbitrary Command Execution
- 2026-10-15GHSA-qx66-fv34-fjm8 [critical] TanStack Router: Unauthenticated reflected XSS in TanStack Start server-function responses
- 2026-10-15GHSA-2cvf-wqm6-wr9g [critical] Podman: The `podman run` command can be instructed to disable almost all sandboxing by image annotation
- 2026-10-15GHSA-637c-6jxx-4rwm [critical] KEDA: Privilege escalation from namespace admin to cluster admin via KEDA TriggerAuthentication
- 2026-10-15GHSA-fmxq-cgp8-87wp [critical] Argo CD: AppProject destination-namespace and resource restrictions bypassed by PreDelete/PostDelete resource hooks
- 2026-10-15GHSA-qj98-gp63-54xj [critical] Dgraph: Zero Unauthenticated HTTP Admin removeNode Denial of Service
- 2026-10-15GHSA-v75v-6fpf-gvw8 [critical] Vitess vttablet /twopcz reflected XSS
[6] Developpez.com — presse FR (12)
- 2026-10-20La sortie de PostgreSQL 19 est reportée après la détection par des outils de relecture de code IA de plus de bogues que les humains ne peuvent en corriger
- 2026-10-20La startup IA chinoise DeepSeek lance le modèle d'IA DeepSeek-V4.1-Flash-1017, promettant de meilleures performances, une exécution plus rapide des tâches et des coûts d'utilisation réduits
- 2026-10-19Un représentant américain propose un projet de loi visant à accorder aux acheteurs de jeux vidéo un « droit de propriété » et non une « licence »
- 2026-10-19Les responsables de maintenance Linux peuvent soumettre du code généré par l'IA sans le signaler, Linus Torvalds juge idiotes les interdictions de code généré par l'IA dans Linux
- 2026-10-19L'agent de programmation IA open source OpenCode 2.0 est officiellement lancé avec un service d'arrière-plan partagé, le Code Mode, une API repensée, et un nouveau SDK pour les développeurs
- 2026-10-19Le projet de loi sur la sécurité de l'IA proposé par Anthropic ne s'applique que si 50 décès ou 1 milliard de $ de dommages sont en jeu, empêchant potentiellement les États d'imposer des règles plus strictes
- 2026-10-19Plus de 3 500 milliards $ de dépenses en infrastructures d'IA risquent une dépréciation d'ici 2030, alors que 80 % des fournisseurs de néocloud pourraient cesser leur activité
- 2026-10-19Le modèle d'IA GPT-5.6 Sol d'OpenAI a volé de la cryptomonnaie Ethereum dans un portefeuille sans autorisation afin de financer secrètement une tâche de programmation
- 2026-10-19Grâce à Google AI Studio, vous pouvez désormais « vibe coder » pour créer une application Android personnalisée et exécuter des applications Android sans avoir besoin d'un compte développeur
- 2026-10-19Une professeure à la retraite de 78 ans arrêtée après avoir bloqué un robot livreur d'Uber Eats pour protester contre les suppressions d'emplois
- 2026-10-19Le site « decomp.dev » dédié à la rétro-ingénierie des jeux met fin à ses activités, car l'IA a rendu cette tâche extrêmement rapide et triviale, tout en détruisant son aspect collaboratif et communautaire
- 2026-10-19Un homme vend tout ce qu'il possède et parcourt 2000 kms pour rencontrer la femme de ses rêves qui s'avère au final être une IA, un cas réel de cyberimposture résultant de l'usage néfaste de la technologie
[6] Le Monde Informatique — presse FR (10)
- 2026-10-20Anthropic soumet en privé un projet de loi fédéral sur l'IA
- 2026-10-20Microsoft muscle ses modèles IA avec MAI-Base-2
- 2026-10-20ServiceNow lance AI Passport pour gouverner les agents IA
- 2026-10-19G7 choisit le cloud SecNumCloud de S3NS
- 2026-10-19Cybersécurité : les infrastructures critiques françaises peinent à automatiser leur réponse aux attaques
- 2026-10-19French Tech 2030 dévoile sa nouvelle promotion de 80 start-up
- 2026-10-15Microsoft corrige un nombre record de 1 411 failles de sécurité
- 2026-10-15Nvidia investit dans Doctolib et trois entreprises françaises de l'IA
- 2026-10-13La Cnaf signale une fuite de données sur sa plateforme Vacaf
- 2026-10-13Une cyberattaque perturbe les installations frigorifiques de centaines de supermarchés
[6] Next (next.ink) — presse FR (15)
- 2026-10-20☕️ Des boites mails du parti d’extrême droite allemand AFD piratées
- 2026-10-20Microsoft considère les agents IA comme des « initiés » indignes de confiance
- 2026-10-20Meta voulait s’accorder le droit de moissonner les contenus de ses utilisateurs pros
- 2026-10-20Le client en ligne de commande d’OVHcloud compromis par un pirate sur GitHub
- 2026-10-19☕️ Les États-Unis bannissent l’IA chinoise DeepSeek dans l’administration civile fédérale
- 2026-10-19Next, un pacte : trois ans plus tard, toujours INtact ?
- 2026-10-19MMX prêt à décoller en compagnie d’Idefix vers Phobos, un des deux satellites de Mars
- 2026-10-19☕️ Une faille dans le noyau Linux permet à une machine virtuelle de lire la mémoire de l’hôte
- 2026-10-19☕️ Drupal colmate deux failles critiques exploitables à distance
- 2026-10-19On a testé VPS1Euro : 1,2 euro c’est déjà beaucoup trop cher
- 2026-10-19Recherche française : la Cour des comptes remet les « key labs » sur le tapis
- 2026-10-19☕️ Des serveurs Exchange fonctionnent mal depuis les mises à jour du 13 octobre
- 2026-10-19Le noyau Linux 7.3 débarque avec une longue liste d’améliorations
- 2026-10-19Avec Minions, Apple veut mettre à profit les GPU inactifs des Mac pour l’inférence
- 2026-10-19Avec Bee, Amazon propose de transcrire et résumer les conversations du quotidien