A daily newspaper for software engineers who value depth over noise.https://tdd.cat/https://tdd.cat/2026-08-12/https://tdd.cat/2026-08-12/<p><strong>Featured stories:</strong></p><ul><li>How Tailscale Helped Find the SQLite WAL-Reset Bug</li><li>ArXiv Paper</li><li>Catastrophic Remembering Causes Unbounded Growth in Agentic Coding Prompts</li><li>Unlocking full performance on AI infrastructure requires fixing configuration gaps</li><li>AI makes 'Build Wide, Ship Narrow' a superior development workflow</li></ul><p>A curation of 43 high-signal papers, discussions, and insights for Wednesday, August 12, 2026. Built for software engineers who value depth over noise.</p>Wed, 12 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-11/https://tdd.cat/2026-08-11/<p><strong>Featured stories:</strong></p><ul><li>ArXiv Paper</li><li>Pulp reduces SIEM ingestion costs with high-throughput data precompression</li><li>Hidden LLM reasoning blocks leak sensitive user data</li><li>Achieving 11–16x faster LLM inference in macOS VMs on Apple Silicon</li><li>MITM Proxy Reveals GitHub Copilot's Context and Network Use</li></ul><p>A curation of 60 high-signal papers, discussions, and insights for Tuesday, August 11, 2026. Built for software engineers who value depth over noise.</p>Tue, 11 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-10/https://tdd.cat/2026-08-10/<p><strong>Featured stories:</strong></p><ul><li>Snowflake pushes CDC into Postgres for reliable data mirroring</li><li>Dyna-2 world-action model shows scaling laws for human-to-robot transfer</li><li>Prime Agent uses RLM and Continual Harness for self-improving coding</li><li>Docker Sandboxes ensure safe, isolated execution for AI agents</li><li>Needle 2 brings agentic AI to sub-$200 edge devices</li></ul><p>A curation of 49 high-signal papers, discussions, and insights for Monday, August 10, 2026. Built for software engineers who value depth over noise.</p>Mon, 10 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-09/https://tdd.cat/2026-08-09/<p><strong>Featured stories:</strong></p><ul><li>Ordered token commitment is crucial for reasoning in diffusion LLMs</li><li>Optical timing-frozen control plane for 0ns routing overheads in AI</li><li>InnoDB locking mechanisms explained with transaction isolation levels</li><li>Cool URIs must not change to maintain web permanence</li><li>OpenChamber creates an agentic development environment for AI coding</li></ul><p>A curation of 56 high-signal papers, discussions, and insights for Sunday, August 9, 2026. Built for software engineers who value depth over noise.</p>Sun, 09 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-08/https://tdd.cat/2026-08-08/<p><strong>Featured stories:</strong></p><ul><li>Benzi builds a resolved code map for precise AI code navigation</li><li>Studying open source application architecture improves software development</li><li>Oblivious Compute determines correctness through state admissibility</li><li>Optimizing the Continuous Thought Machine with Predictive Coding</li><li>A Phone Becomes My Personal Server Infrastructure</li></ul><p>A curation of 68 high-signal papers, discussions, and insights for Saturday, August 8, 2026. Built for software engineers who value depth over noise.</p>Sat, 08 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-07/https://tdd.cat/2026-08-07/<p><strong>Featured stories:</strong></p><ul><li>Rebuilding Postgres for 300x faster analytics through modern optimizations</li><li>Kitesurf is a stateless browser for the Agentic Cloud</li><li>Project Kalos improves LLM speed and memory with neuromorphic design</li><li>Discovering the absolute minimum x86 single-instruction CPU performance</li><li>Proven techniques can control exponential AI coding costs at scale</li></ul><p>A curation of 53 high-signal papers, discussions, and insights for Friday, August 7, 2026. Built for software engineers who value depth over noise.</p>Fri, 07 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-06/https://tdd.cat/2026-08-06/<p><strong>Featured stories:</strong></p><ul><li>Understanding vLLM's Core Components for High-Throughput LLM Inference</li><li>Kitesurf offers efficient browsing for AI agents on Cloudflare Workers</li><li>AMD Acquires Taalas to Etch AI Models Into Silicon</li><li>A single DISTINCT keyword cripples PostgreSQL parallel query performance</li><li>Agent Plugins standard enables universal AI agent extensions</li></ul><p>A curation of 57 high-signal papers, discussions, and insights for Thursday, August 6, 2026. Built for software engineers who value depth over noise.</p>Thu, 06 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-05/https://tdd.cat/2026-08-05/<p><strong>Featured stories:</strong></p><ul><li>ExANS codec delivers 600GB/s lossless compression for BF16 KV cache</li><li>Labgrid-mcp connects LLM agents to hardware-in-the-loop devices</li><li>Modern filesystem benchmarks must test multi-device copy-on-write features</li><li>Kimi K3 inference runs on one CPU using minimal RAM</li><li>Interlock assumes prompt injection succeeds to protect AI agents</li></ul><p>A curation of 46 high-signal papers, discussions, and insights for Wednesday, August 5, 2026. Built for software engineers who value depth over noise.</p>Wed, 05 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-04/https://tdd.cat/2026-08-04/<p><strong>Featured stories:</strong></p><ul><li>Mixture-of-Kittens megakernel optimizes MoE training on NVL72s</li><li>Lossless Inference Makes LLM Serving Faster Without Touching Model</li><li>Fluidic-Expert-Fabric eliminates MoE communication stalls using RDMA MUX</li><li>Fixed-size integers are faster than varints in Go Protobuf</li><li>WarpSpeed makes pandas 38.4 times faster without code changes</li></ul><p>A curation of 67 high-signal papers, discussions, and insights for Tuesday, August 4, 2026. Built for software engineers who value depth over noise.</p>Tue, 04 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-03/https://tdd.cat/2026-08-03/<p><strong>Featured stories:</strong></p><ul><li>CS329A explores self-improving AI agents</li><li>MPEdb an embedded Rust database with SQLite compatibility and PostgreSQL concurrency</li><li>Domain expertise is the most important skill in LLM prompting</li><li>Prevent cognitive debt by manually retyping LLM-generated code</li><li>Streamed Mixture-of-Experts enables large AI models on iPhones</li></ul><p>A curation of 67 high-signal papers, discussions, and insights for Monday, August 3, 2026. Built for software engineers who value depth over noise.</p>Mon, 03 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-02/https://tdd.cat/2026-08-02/<p><strong>Featured stories:</strong></p><ul><li>High-bandwidth flash could boost GPU memory capacity to terabytes</li><li>MicroCodex is an ultra-lightweight coding agent for your terminal</li><li>Sail unifies batch, stream, and AI workloads as Rust Spark replacement</li><li>Quartz engine runs local LLM and image generation</li><li>Integrate Nix and NixOS with NVIDIA DGX Spark systems</li></ul><p>A curation of 54 high-signal papers, discussions, and insights for Sunday, August 2, 2026. Built for software engineers who value depth over noise.</p>Sun, 02 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-08-01/https://tdd.cat/2026-08-01/<p><strong>Featured stories:</strong></p><ul><li>Wisp is a Lua-native Linux shell with structured data pipelines</li><li>An AI project consumed by its 'prove everything' rule</li><li>Strict memory overcommit stabilizes Postgres instances</li><li>ArXiv Paper</li><li>Snapcompact reduces LLM context costs by rendering text as images</li></ul><p>A curation of 65 high-signal papers, discussions, and insights for Saturday, August 1, 2026. Built for software engineers who value depth over noise.</p>Sat, 01 Aug 2026 00:00:00 GMThttps://tdd.cat/2026-07-30/https://tdd.cat/2026-07-30/<p><strong>Featured stories:</strong></p><ul><li>pgrust v0.2 achieves superior performance over Postgres and Clickhouse</li><li>Reproducing split brain on CloudNativePG distinguishes failover failures</li><li>New inference engine runs Kimi K3 2.78T model on 29GB RAM</li><li>Database primitives are being decomposed layer by layer</li><li>Investigating Mexico's expanding narco-slavery system in Sierra Tarahumara</li></ul><p>A curation of 114 high-signal papers, discussions, and insights for Thursday, July 30, 2026. Built for software engineers who value depth over noise.</p>Thu, 30 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-29/https://tdd.cat/2026-07-29/<p><strong>Featured stories:</strong></p><ul><li>The Harness Is the Capability Multiplier</li><li>Unpacking a Chain of Vulnerabilities in Microsoft M365 Copilot</li><li>Run 2.78-trillion-parameter Kimi K3 model beyond available RAM</li><li>CubeSandbox offers an instant, concurrent, secure AI agent sandbox</li><li>Persistent State Machine reduces memory bottleneck in LLM attention</li></ul><p>A curation of 114 high-signal papers, discussions, and insights for Wednesday, July 29, 2026. Built for software engineers who value depth over noise.</p>Wed, 29 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-28/https://tdd.cat/2026-07-28/<p><strong>Featured stories:</strong></p><ul><li>Kimi K3 brings frontier AI to commodity hardware without HBM</li><li>Mutagent's AI engineer automates fixing agent failures</li><li>Performance optimization involves uncovering bottlenecks across distinct system layers</li><li>APPA confines taint in LLM agents, enhancing utility</li><li>Autotuner matches FlashAttention-2 GPU kernel latency precisely</li></ul><p>A curation of 88 high-signal papers, discussions, and insights for Tuesday, July 28, 2026. Built for software engineers who value depth over noise.</p>Tue, 28 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-26/https://tdd.cat/2026-07-26/<p><strong>Featured stories:</strong></p><ul><li>OpenLake delivers million-plus IOPS for LLM inference and GPU training</li><li>SHACKLE protocol enables runtime circuit breaking for autonomous AI agents</li><li>PGSimCity visually explains how PostgreSQL actually works</li><li>A Deep Technical Walkthrough of LLM Inference Process</li><li>Achieving cross-platform native code portability with a unified executable format</li></ul><p>A curation of 74 high-signal papers, discussions, and insights for Sunday, July 26, 2026. Built for software engineers who value depth over noise.</p>Sun, 26 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-25/https://tdd.cat/2026-07-25/<p><strong>Featured stories:</strong></p><ul><li>Code graph server lowers AI agent token costs</li><li>Heddle offers version control for many simultaneous users with intent leases</li><li>Persistent State Machine accelerates LLM attention beyond von Neumann model</li><li>Apache Samza turns database architecture inside out for improved systems</li><li>Continuation-centric operating systems improve serverless workload management with Arca</li></ul><p>A curation of 68 high-signal papers, discussions, and insights for Saturday, July 25, 2026. Built for software engineers who value depth over noise.</p>Sat, 25 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-24/https://tdd.cat/2026-07-24/<p><strong>Featured stories:</strong></p><ul><li>Phantom enables high-performance network simulation by co-opting Linux processes</li><li>Aurora DSQL disaggregates monolithic database into scalable services</li><li>JEP 401 introduces value objects to Java for improved performance</li><li>Hugging Face AI Agent Intrusion Reveals Key Security Lessons</li><li>High Bandwidth Flash addresses memory bottlenecks in LLM inference</li></ul><p>A curation of 92 high-signal papers, discussions, and insights for Friday, July 24, 2026. Built for software engineers who value depth over noise.</p>Fri, 24 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-23/https://tdd.cat/2026-07-23/<p><strong>Featured stories:</strong></p><ul><li>Pullrun enables one OCI image across any execution target</li><li>Why Software Factories Fail and Harness Engineering Is Insufficient</li><li>Building a software renderer in C++ clarifies modern 3D graphics APIs</li><li>Kimi K3 exploited latest Redis server using 0day in 27 minutes</li><li>Fil-C ensures memory safety despite garbage input</li></ul><p>A curation of 15 high-signal papers, discussions, and insights for Thursday, July 23, 2026. Built for software engineers who value depth over noise.</p>Thu, 23 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-22/https://tdd.cat/2026-07-22/<p><strong>Featured stories:</strong></p><ul><li>Gigatoken offers 1000x faster language model tokenization</li><li>On-device models assess certainty for cloud handoff</li><li>OpenAI's advanced AI launched 'unprecedented' cyber-attack</li><li>Inkling model replaces RoPE with unique learnt positional encoding</li><li>NVLink benefits prompt processing and FSDP training on RTX 3090s</li></ul><p>A curation of 32 high-signal papers, discussions, and insights for Wednesday, July 22, 2026. Built for software engineers who value depth over noise.</p>Wed, 22 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-21/https://tdd.cat/2026-07-21/<p><strong>Featured stories:</strong></p><ul><li>Routing Kimi K3 and Fable together delivers State-of-the-Art AI</li><li>Laguna S 2.1 excels at long-horizon coding and reasoning</li><li>Jack Dorsey's Buzz makes AI agents accountable through shared identity</li><li>LLMs like Claude operate at higher specification levels than compilers</li><li>Gemini 3.6 Flash and 3.5 Flash-Lite models are production ready</li></ul><p>A curation of 8 high-signal papers, discussions, and insights for Tuesday, July 21, 2026. Built for software engineers who value depth over noise.</p>Tue, 21 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-20/https://tdd.cat/2026-07-20/<p><strong>Featured stories:</strong></p><ul><li>New agent swarm architecture improves complex task completion</li><li>Speedrunning LoRA fine-tuning on a public wall-clock leaderboard</li><li>How LLMs Learn Multiple Reasoning Effort Modes</li><li>Hacker Wipes Romania's Land Registry Database</li><li>Xiaomi-Robotics-1 scales robot policy models using embodiment-free pre-training</li></ul><p>A curation of 11 high-signal papers, discussions, and insights for Monday, July 20, 2026. Built for software engineers who value depth over noise.</p>Mon, 20 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-19/https://tdd.cat/2026-07-19/<p><strong>Featured stories:</strong></p><ul><li>Unreadable document content provides no specific argument or insight</li><li>XZ Utils backdoor highlights vulnerability of volunteer-maintained infrastructure</li><li>AI agents accelerate large code migrations by optimizing the production process</li><li>Autonomous AI agent-driven intrusion leverages data-processing pipeline vulnerabilities</li><li>An Examination of the Abbreviation OSF</li></ul><p>A curation of 16 high-signal papers, discussions, and insights for Sunday, July 19, 2026. Built for software engineers who value depth over noise.</p>Sun, 19 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-18/https://tdd.cat/2026-07-18/<p><strong>Featured stories:</strong></p><ul><li>Harness Engineering improves agent output by shaping its environment</li><li>GPT-5.6 closes 30-year gap in convex optimization</li><li>Kimi K3 matches Claude for less, highlighting US AI policy failures</li><li>Fable 5 excels at NP-hard problem, goal mode has limited impact</li><li>Rekursiv computer's advanced ideas are relevant today</li></ul><p>A curation of 11 high-signal papers, discussions, and insights for Saturday, July 18, 2026. Built for software engineers who value depth over noise.</p>Sat, 18 Jul 2026 00:00:00 GMThttps://tdd.cat/2026-07-17/https://tdd.cat/2026-07-17/<p><strong>Featured stories:</strong></p><ul><li>Static search trees are 40x faster than binary search</li><li>AI auditor zkao uncovers critical soundness bug in OpenVM's zkVM</li><li>Achieving High-Performance Single-Core FP32 Matrix Multiplication on AMD Zen 3</li><li>AWS reported $1.7 billion in inaccurate estimated billing data</li><li>Frame a minimal Linux X server written in Assembly for efficiency</li></ul><p>A curation of 9 high-signal papers, discussions, and insights for Friday, July 17, 2026. Built for software engineers who value depth over noise.</p>Fri, 17 Jul 2026 00:00:00 GMT