181
Making Postgres slower
As an respectable unemployed person must do, I tried to make Postgres as slow as possible
New
The latest blog posts to take off on Hacker News. 1,888 posts.
182
PSA: SQLite WAL checksums fail silently and may lose data
SQLite WAL has checksums, but on corruption it drops all the data and does not raise error
183
Async I/O on Linux in databases
I've been working on a complex multi-model database for a few weeks now, and recently I took time to simplify and test out an idea I had on a simple key-value database. I started with the basics: A hash table in memory, a simple append-only log
184
Parsing Protobuf like never before
Historically I have worked on many projects related to high-performance Protobuf, be that on the C++ runtime, on the Rust runtime, or on integrating UPB , the fastest Protobuf runtime, written by my colleague Josh Haberman . I generally don’t post directly about my current job, but my most recent toy-turned-product is…
185
Extending That XOR Trick to Billions of Rows
Learn how to extend the classic XOR trick to find thousands of missing values using Invertible Bloom Filters
186
Why LLMs Can't Write Q/Kdb+: Writing Code Right-to-Left
In my previous post I showed how LLMs struggle with coding in q/kdb+; having immediate feedback loops using the q-MCP server does help a…
187
Mini NASes marry NVMe to Intel's efficient chip
I'm in the process of rebuilding my homelab from the ground up, moving from a 24U full-size 4-post rack to a mini rack. One of the most difficult devices to downsize (especially economically) is a NAS. But as my needs have changed, I'm bucking the trend of all datahoarders and I need less storage than the 120 TB (80…
188
Data on AI-related Show HN posts
The idea for this article didn't exist in my brain before this morning. But there I was, scrolling the New page and even more tired of all the AI-related Show HN posts than usual. I was confident that their numbers were multiplying and wanted proof. Exactly how much more AI crap is on my lawn compared to last year?
189
Building agents using streaming SQL queries
Table of Contents Agents Need to Interact With LLMs Agents Should Be Event-Driven Agents Need Context Agents Require Memory When SQL Is Not Enough Parting Thoughts AI Agents have improved in leaps …
190
Wrong ways to use the databases, when the pendulum swung too far
Disclaimer: This story took place when I was a junior developer for a previous employer ages ago. Some details might be fuzzy, and goes without saying this is entirely from my own POV so take everything here with humor and a grain of salt.
191
I made a search engine worse than Elasticsearch (2024)
Integrating my BM25 pandas search library, SearchArray, into BEIR, in order to embarass myself in public.
192
How to Store Data on Paper?
How to print my voice? How to put music on paper? How to print an executable program? All this boils down to storing digital data on paper. This is also called paper data storage . Arrived in this domain from poetry, I’ve been investigating the area for some time. Here is what I found.
193
Just make it scale: An Aurora DSQL story
AWS Senior Principal Engineers, Niko Matsakis and Marc Bowes, take us inside Aurora DSQL's development: scaling write operations without two-phase commit, overcoming garbage collection hurdles, and embracing Rust for both data and control planes.
194
Loading Pydantic models from JSON without running out of memory
Pydantic’s JSON loading uses a huge amount of memory; here’s how to reduce it.
195
A simple search engine from scratch
*if you include word2vec.
196
197
The principles of database design, or, the Truth is out there
Every software project needs to represent the reality of the business he is embedded in. The way we can represent reality as limited rational beings is through propositions, i.e, declarative statements that affirm or deny something about reality. When a collection of such propositions is stored in a computer system…
198
199
21 GB/s CSV Parsing Using SIMD on AMD 9950X
Sep 0.10.0 was released April 22nd, 2025 with optimizations for AVX-512 capable CPUs like the AMD 9950X (Zen 5) and updated benchmarks including the 9950X. Sep now achieves a staggering 21 GB/s on the 9950X for the low-level CSV parsing. 🚀 Before 0.10.0, Sep achieved ~18 GB/s on 9950X.
200
DuckDB is probably the most important geospatial software of the last decade
DuckDB’s spatial extension cut geo setup to two lines of SQL, and interest in geospatial data turned upward the moment it shipped.
201
Bloom Filters
The original motivation for the creation of Bloom filters is efficient set membership, using a probabilistic approach to significantly reduce the time and space required to reject items that are not members in a certain set.
202
Derivation and Intuition behind Poisson distribution
Made in Notion, the connected workspace where AI agents and teams organize docs, projects and knowledge together.
203
A faster way to copy SQLite databases between computers
Dumping a SQLite database to a text file can make it much smaller, which means you can download it faster.
204
Reverse geocoding is hard
My wife and I run OpenBenches - a crowd-sourced database of nearly 40,000 memorial benches. Every bench is geo-tagged with a latitude and longitude. But how do you go from a string of digits to something human readable? How do I turn -33.755780,150.603769 into "42 Wallaby Way, Sydney, Australia"? Luckily, that's a…
205
Shardines: SQLite3 Database-per-Tenant with ActiveRecord
There is a pattern I am very fond of - “one database per tenant” in web applications with multiple, isolated users. Recently, I needed to fix an application I had for a long time where this database-per-tenant multitenancy utterly broke down, because I was doing connection management wrong. Which begat the question…
206
Street address errors in Google Maps
(or actually they do, but they don’t use this knowledge effectively)Update, April 26, 2025: the address fix for W 6th Ave is live, mostly. Going forward I wish that Google Maps would make it harder…
207
Decomposing Transactional Systems
Every transactional system must execute, order, validate, and persist transactions.
208
Dumb statistical models, always making people look bad
This is Jessica. In my Prediction for Decision-making class last quarter we talked about human predictions and decisions relative to algorithmic ones. This led to a discussion about why it’s often hard to demonstrate the value of human knowledge once you have a decent statistical model.
209
I analyzed chord progressions in 680k songs
And the results surprised me
210
SQLite transactions and virtual tables
In my previous post I introduced SQL virtual tables and how to use them in Go—registering modules per connection, defining schemas, and querying external sources as regular tables.