781
Optimizing a WebGPU Matmul Kernel for 1 TFLOP
Building Surfgrad, a high-performant, WebGPU-powered autograd library
New
The latest blog posts to take off on Hacker News. 4,387 posts.
782
Way too many ways to wait on a child process with a timeout
I often need to launch a program in the terminal in a retry loop. Maybe because it's flaky, or because it tries to contact a remote service that is not available. A few scenarios:
783
The number given as % CPU in Activity Monitor
One of the most common reasons for opening Activity Monitor is to check the % CPU of processes that might be running out of control. How accurate are those figures, though, and should you believe t…
784
We built a self-healing system to survive a concurrency bug at Netflix
Our CPUs were dying, the bug was temporarily un-fixable, and we had no viable path forward. Here's how we managed to survive.
785
Useful built-in macOS command-line utilities
macOS comes with a lot of built-in utilities. Here's a list of some that I find interesting.
786
Traceroute Isn't Real
Almost nobody understands how traceroute works. Worse, it's not a real tool
787
Why systemd is a problem for embedded Linux
I don’t want to post yet another anti-systemd rant, but I think the ubiquity of systemd could eventually create real problems for embedded Linux. I just wish I knew what to do about it.
788
789
The Dual Nature of Events in Event-Driven Architecture
On Software Architecture, Programming Languages,…2024112 points63 commentsSoftware Engineering details
Given that events play such a central role in event-driven architecture, there’s an astonishing lack of agreement on what should be contained in an event. This may be rooted in the fact that, depending on your perspective, events fulfill different purposes.
790
OpenZFS deduplication is good now and you shouldn't use it
OpenZFS 2.3.0 will be released any day now, and it includes the new “Fast Dedup” feature. My team at Klara spent many months in 2023 and 2024 working on it, and we reckon it’s pretty good, a huge step up from the old dedup as well as being a solid base for further improvements. I’ve been watching various forums and…
791
Eighteen Years of ABI Stability
It has been eighteen years of libcurl ABI stability.
792
793
Improving Xwayland window resizing
One of the quickest ways to determine whether particular application runs using Xwayland is to resize one of its windows and see how it behaves, for example
794
795
NixOS is not reproducible
Okay, sorry for the clickbait. NixOS is not reproducible according to the Reproducible Builds definition. I keep reading people making this claim repeatedly on orange-site, even LWN.net made a similar claim when writing about Nix and Guix earlier this week.1 Along with their recently launched wiki. So, what is the…
796
Zero or Sign Extend
A while back I had to deal with a bit-packed format that contained a list of integer values encoded in one of a pre-defined sets of bit widths, where both the allowed bit widths and the signed-ness…
797
RISC-V is currently slow compared to modern CPUs
RISC-V is currently quite slow compared to modern CPUs in 2024.
798
Optimizers need a rethink
Thoughts on how compiler and database optimizers should evolve in the future, focusing on predictability and guarantees instead of smarts and heroics
799
Against /tmp
I commented on Lobsters that /tmp is usually a bad idea , which caused some surprise. I suppose /tmp security bugs were common in the 1990s when I was learning Unix, but they are pretty rare now so I can see why less grizzled hackers might not be familiar with the problems.
800
MVCC – the part of PostgreSQL we hate the most (2023)
As much as Andy loves PostgreSQL, there is one part that is terrible and causes many headaches for people. Learn what it is and why it sucks.
801
Blocking code is a leaky abstraction
Asynchronous code does not require the rest of your code to be asynchronous. I can’t say the same for blocking code.
802
803
Zero-latency SQLite storage in every Durable Object
Kenton Varda introduces the next iteration of Cloudflare's Durable Object platform, which recently upgraded from a key/value store to a full relational system based on SQLite. For useful background on …
804
Lessons learned from profiling an algorithm in Rust
Recently, I've been working on a new approximate nearest neighbor search algorithm called RaBitQ . The author has already provided a C++ implementation that runs quite fast. I tried to rewrite it in Rust (yet another RiiR). But I found that my implementation is much slower than the original one. Here is how I improve…
805
Node.js, Pipes, and Disappearing Bytes
Someone in one of my online communities recently asked about why their Node command was dropping output when piped to another command:
806
Every bug/quirk of the Windows resource compiler (rc.exe), probably
Fuzz testing decades-old software can turn up some curious behaviors
807
Designing a Fast Concurrent Hash Table
I recently released papaya, a fast and feature-complete concurrent hash table for Rust. In this post I want to dive into the design and research that went into creating it, as well as why you might consider using it over existing solutions.
808
WASM Is the New CGI
I read a wonderful twitter thead about CGI and the birth of the web and this triggered a thought I've been kicking around. > Wasm is the new CGI And to be cle...
809
Practices of Reliable Software Design
It should have good performance and be able to hold many entries. Reads are more common than writes. I know how I would do it already, but I’m curious about your approach.
810
The costs of the i386 to x86-64 upgrade
If you read my previous article on DOS memory models, you may have dismissed everything I wrote as “legacy cruft from the 1990s that nobody cares about any longer”. It's time to see how any of that carried over through the 16-bit to 64-bit evolution.