El Solitario

  • GNU/Linux
  • Tutorials
  • Opinion

Because being a geek is a good thing

Diagram of JWT tokens split into header, payload, and signature
Programación

JWT: How the Tokens That Replace Server-Side Sessions Work

A JWT token lets a server confirm who you are without querying the database again on every request. This guide explains its structure, how to sign it with Node.js, and its real security risks.

By Andrés Morales, 1 month29 July, 2026 ago
AI agents reviewing an extensive corporate handbook on a screen
Inteligencia Artificial

HANDBOOK.md: AI agents fail 64% of tasks with long policies

HANDBOOK.md is a new benchmark that measures whether AI agents follow lengthy corporate handbooks during long tasks. The best of 30 configurations evaluated passes just 36.2% of the tests under strict grading.

By Andrés Morales, 1 month29 July, 2026 ago
Jailbroken Kindle running Tailscale in proxy and TUN mode
Noticias Tech

Tailscale Adds Proxy and TUN Mode to Jailbroken Kindles

A community fork of the Tailscale client for jailbroken Kindle adds proxy mode with SOCKS5 and HTTP CONNECT, plus full TUN mode on some models. KOReader can now reach Calibre, Wallabag, and Audiobookshelf servers over the tailnet, and the same plugin works on Kobo and PocketBook.

By Andrés Morales, 1 month29 July, 2026 ago
Illustration of an own website syndicating content to Substack through the POSSE method
Noticias Tech

IndieWeb and POSSE: Why a Substack Blog Doesn’t Replace Your Own Website

An essay published on June 10, 2026 by blogger Elizabeth Tai argues that Substack is a distribution tool, not an own website. The alternative she proposes, known as POSSE, reverses the publishing order: your own domain first, then syndication to Substack, Mastodon or X.

By Andrés Morales, 1 month28 July, 2026 ago
Diagram of the Kimi Delta Attention state updating token by token
Noticias Tech

Kimi Delta Attention: The Linear Attention Behind Kimi K2 and Qwen3

Kimi Delta Attention (KDA) is the linear attention component that Moonshot AI and Qwen3 use to process long context without softmax’s quadratic cost. It comes from a family that started with DeltaNet and went through Gated DeltaNet, adding increasingly fine control over what the model remembers.

By Andrés Morales, 1 month28 July, 2026 ago
Conceptual illustration of open-weights models and AI computing chips
Inteligencia Artificial

Anthropic Clarifies It Never Called for Banning Open-Weights Models

Dario Amodei published a statement on July 27, 2026 denying that Anthropic wants to ban open-weights models, including Chinese ones. Instead, he proposes restricting chip sales to China, targeting industrial-scale distillation, and requiring mandatory safety testing for every capable model.

By Andrés Morales, 1 month27 July, 2026 ago
Flags of China and the United States symbolizing the AI hegemonism dispute over Moonshot AI
Inteligencia Artificial

China Accuses US of AI Hegemonism Over Moonshot AI Investigation

China accused the United States of practicing AI hegemonism after Washington threatened to investigate Moonshot AI, the startup behind the Kimi models. Nvidia CEO Jensen Huang defended the quality of Chinese models and called for setting aside unfounded fears.

By Andrés Morales, 1 month27 July, 2026 ago
Diagram of the TLS 1.3 handshake between client and server
Seguridad

TLS 1.3: How the Handshake That Encrypts HTTPS Works

The TLS 1.3 handshake reduced to a single round trip the negotiation that used to take two in TLS 1.2, and eliminated cipher suites without forward secrecy. Here we explain how it works step by step, with real commands to inspect and configure it in production.

By Clara Vásquez, 1 month27 July, 2026 ago
Destructive scanning machine cutting the spine of rare books for AI
Inteligencia Artificial

Anthropic Scans and Destroys Rare Books to Train Its AI

AI companies buy rare books in bulk through ISBNdb, scan them with machines that cut the spine, and destroy the original. A federal judge ruled the practice fair use, and Anthropic hired Google Books’ former head of partnerships for an internal project known as Project Panama.

By Andrés Morales, 1 month27 July, 2026 ago
Screen showing Lean code and formal proofs generated by an LLM
Noticias Tech

LLMs Test Code in Lean: How a zstd Decoder Was Verified

A Google engineer implemented a Zstandard decompressor in Lean with formal proofs generated by an LLM, motivated by the historically steep cost of verifying software like seL4. The experiment explores whether artificial intelligence can make practical an approach that until now was territory reserved for a few teams with nearly unlimited budgets.

By Andrés Morales, 1 month26 July, 2026 ago

Posts pagination

Previous 1 … 14 15 16 … 22 Next
Linux Courses
Introducción a Linux: Instala Linux en tu PC

Introducción a Linux: Instala Linux en tu PC

Empieza seguro en Linux

4.2 | Gratis


Aprendiendo Linux: uso, terminal y comandos

Aprendiendo Linux: uso, terminal y comandos

Aprende sobre el entorno de escritorio y los comandos básicos de GNU/Linux

4.7 | $19.99

El Solitario
Recent Posts
  • 17,000 Words Hand-Justified: rs1n’s Super Metroid Guide
  • Haiku R1/beta6 Is Now Available, Two Years After Beta5
  • Kernel.org Burns 14 CPU Cores Tracking AI Crawlers
  • AI Guides Live Removal of a Brain Tumor During Surgery
  • Erasure Coding: How HDFS Cuts Replication Cost from 200% to 50%
Categories
  • Inteligencia Artificial
  • Microservices
  • Noticias Tech
  • Programación
  • Redes
  • Seguridad
  • Tutorials
Recent Comments
  • Luis Navarro on Jacobian Conjecture: 87 Years, and the Counterexample Came From AI (in Dimension 3)
  • Trevor Uptain on Jacobian Conjecture: 87 Years, and the Counterexample Came From AI (in Dimension 3)
  • Privacy Policy
Hestia | Developed by ThemeIsle