New
The latest blog posts to take off on Hacker News. 49,212 posts.
We overestimate our short-term ability, but underestimate our long-term ability
My way of thinking about long-term habit building
Clog: Making a homegrown ClickHouse log for $20 a month
There was recently a discussion on Hacker News around application logging on a budget. At work I've been trying to keep things lean, not to the point of absurdity, but also not using a $100 or $1000/month setup, when a $10 one will suffice for now . We settled on a homegrown Clickhouse + PHP solution that has…
<3 Deno
Deno is a relatively new JavaScript runtime. I find it quite interesting and aesthetically appealing, in-line with the recent trend to rein in the worse-is-better law of software evolution. This post explains why.
Why I chose OpenAI over academia
Part 2 of a blog post about my time on the computer science job market in 2021-2022, in which I discuss my personal decision-making process between offers.
Algorithmic Trading: A Practitioner’s Guide
I really enjoyed reading Algorithmic Trading: A Practitioner’s Guide by Jeffrey M. Bacidore. Before starting, I imagined it would cover various strategies for trading in the markets, along th…
How to create portable Linux binaries
Creating application binaries for Linux that run on a wide range of distributions is a bit tricky, as different distributions ship different versions of various system libraries. These are usually backwards compatible, but not forwards compatible, so programs linked against older versions of the libraries also work…
Sun: The Network Is the Computer
The story of Sun Microsystems and the Java programming language
Simple Physics-based Flight Simulation with C++
Creating simple but realistic physics-based flight simulations from scratch with C++ and OpenGL.
Problem solving with dimensional analysis
A dimension is a physical quantity that can be measured, while a unit is a specific amount of a physical quantity. For example, time is a dimension, which can be measured in temporal units such as hours or seconds.
Let's write a setjmp
Yesterday I wrote that setjmp is handy and that it would be nice to have without linking the C standard library. It’s conceptually simple, after all. Today let’s explore some differently-portable implementation possibilities with distinct trade-offs. At the very least it should illuminate why setjmp sometimes requires…
I had ChatGPT write a Decemberists song and recorded it
And I recorded it. Here's "Sailor's Song."
How to train large models on many GPUs? (2021)
[Updated on 2022-03-13: add expert choice routing.] [Updated on 2022-06-10]: Greg and I wrote a shorted and upgraded version of this post, published on OpenAI Blog: “Techniques for Training Large Neural Networks”
My review of the C standard library in practice
In general, when working in C I avoid the standard library, libc, as much as possible. If possible I won’t even link it. For people not used to working and thinking this way, the typical response is confusion. Isn’t that like re-inventing the wheel? For me, libc is a wheel barely worth using — too many deficiencies in…
Creatively Misusing TLA+
Five not-quite-normal uses
In Defense of Not-Invented-Here Syndrome (2001)
Time for a pop quiz. 1. Code Reuse is: a) Goodb) Bad 2. Reinventing the Wheel is: a) Goodb) Bad 3. The Not-Invented-Here Syndrome is: a) Goodb) Bad Of course, everybody knows that you should always…
Treasure trove of mechanical animations (2015)
Whenever I watch videos of how things are made, I marvel at the cleverness of the manufacturing machines. Retired engineer Duc Thang Nguyen has created over 170
Money Laundering and AML Compliance
A brief overview of what money laundering is, what financial institutions do about it, and why.
Understanding and coding the self-attention mechanism of large language models
Step-by-step tutorial explaining scaled dot-product self-attention for large language models, with Python code that builds the mechanism from scratch.
QGIS is the mapping software you didn't know you needed
QGIS is incredible: You can build maps that make use of the absolute wealth of public data out there and put Google Maps to absolute shame. This article is doing just that.
I changed my mind about nuclear waste
I’ve feared it. I’ve protested it. Now I’m hugging it
How a Zig IDE Could Work
Zig is a very interesting language from an IDE point of view. Some aspects of it are friendly to IDEs, like a very minimal and simple-to-parse syntax (Zig can even be correctly lexed line-by-line, very cool!), the absence of syntactic macros, and ability to do a great deal of semantic analysis on a file-by-file basis…
Is Seattle a 15-minute city? It depends on where you want to walk
How many Seattleites can walk to fulfill their daily needs?
Service mesh use cases (2020)
Service mesh is a blazing hot topic in software engineering right now and rightfully so. I think it’s extremely promising technology and I’d love to see it widely adopted (in cases where it truly makes sense). Yet it remains shrouded in mystery for many people and even those who are familiar with it have trouble…
A GPT in 60 Lines of NumPy
Implementing a GPT model from scratch in NumPy.