121
Ncdu 2: Less hungry and more Ziggy
Ncdu (NCurses Disk Usage) is a terminal-based disk usage analyzer for Linux and other POSIX-y systems, (formerly) written in C and available in the package repositories of most distributions.
Topics /
Systems & Performance
The most popular blog posts about systems & performance on Hacker News. 358 posts.
122
WinGet is terrible, I want AppGet back
It’s been almost a year that Keivan Beigi shut down the AppGet project in response to shenanigans by the Windows App Model team. Microsoft…
123
Urbit: The good, the bad, and the insane
A whirlwind tour through the bucket-load of fail that is Urbit.
124
Teaching a cheap ethernet switch new tricks (2019)
Ethernet rules everything around us, a large proportion of our systems communicate to each other with ethernet somewhere in the line. And the fast pace race to the bottom for embedded systems means that almost all network equipment is smart to some degree these days.
125
DNS doesn't “propagate”
DNS "propagation" is actually caches expiring
126
Leap seconds: Causing bugs even when they don’t happen
Some new minor leap second drama is coming up, and for once we can
127
Why isn't differential dataflow more popular?
Differential dataflow is a library that lets you write simple dataflow programs and a) then runs them in parallel and b) efficiently updates the outputs when new inputs arrive. Compared to competition like spark and kafka streams , it can handle more complex computations and provides dramatically better throughput and…
128
2021 GPS Accuracy Issue Impacting Some Garmin, Suunto, other GPS Devices
Update: As of Sunday morn...
129
Linux, macOS, and Windows running simultaneously on a first gen Core i5
This is my main computer, a Thinkpad T410 with a 1st generation Intel Core i5 and 8 GB of RAM. It runs Arch Linux with Xfce.
130
131
Mezzano on Librebooted ThinkPads
I decided to try running Mezzano on real hardware. I figured my Librebooted ThinkPads would be good targets, since, thanks to Coreboot and the Linux kernel, I have reference source code for all the hardware.
132
The many lies about reducing complexity part 2: Cloud
That standard image that explains the difference between on-premises, IAAS, PAAS, and SAAS in the cloud? You know what, it is hugely misleading. Here is the real deal.
133
Hunting down the stuck BGP routes
BGP is the glue between all of the thousands of border routers that make up the internet (you can find this post (battleships) and this post (EvE) as a crash course on how BGP works).
134
PDP-11 Booting
Instructions for booting a PDP-11/34 with RT-11 from a RK05
135
The Linux Distributions of 1992
The modern Linux distros owe their existence to those early releases, back in 1992.
136
M1 Pro First Impressions: Core Management and CPU Performance
The P cores in the M1 Pro/Max CPU are managed in two groups of four, sparing load on the second group, and distributing it unevenly within each group. Its two E cores outperform the four in the M1 …
137
138
Speeding Up the Webcola Graph Viz Library with Rust and WebAssembly
For a recent project I've been working on, I wanted to include a graph showing the relationships between different artists on Spotify. Spotify provides the data directly from their API, and I had everything set up to pull it for a user's top artists and into the browser.
139
Understanding Connections and Pools
What connections are, how they affect our systems, and how and why pooling works.
140
The QOI File Format Specification
Last month I announced a toy project called QOI — the Quite OK Image Format . It losslessly compresses RGB and RGBA images to a similar size of PNG, while offering a 20x-50x speedup in compression and 3x-4x speedup in decompression.
141
Fun with Unix domain sockets
A small enhancement to Datasette this week: I’ve added support for proxying via Unix domain sockets. This started out as a feature request from Aslak Raanes: #1388: Serve using UNIX …
142
PostgreSQL, Memory and the Cloud
This post explains how Linux memory overcommit and OOM-killer behavior can turn a single runaway query into a full PostgreSQL crash, and why many managed cloud database services fail to disable overcommit the way self-hosted PostgreSQL recommends.
143
Optimizations Enabled by -ffast-Math
This blog post describes the optimizations enabled by -ffast-math when compiling C or C++ code with GCC 11 for x86_64 Linux (other languages/operating systems/CPU architectures may enable slightly different optimizations).
144
145
Speculating the entire x86-64 instruction set in seconds with one weird trick
As cheesy as the title sounds, I promise it cannot beat the cheesiness of the technique I’ll be telling you about in this post. The morning I saw Mark …
146
I wrote a linker everyone can understand
When we wrote our assembler we made many design choices that helped us cut down on the complexity of the assembler itself. One of those decisions was to have the assembler directly output executables. That was a good choice in that it allowed us to quickly test as we were developing: once we implemented an opcode, we…
147
Benchmarking the Apple M1 Max
💻 Casino brand Royal reels casino 🎯 Alternative names Royal reels, Royal reels casino Australia, Royal Reels pokies, Royal reels online casino 📅 Launch year 2021 🔗 Royal reels mirrors 17, 18, 19, 21, 22, 23, 24, 26 🌐 Supported regions Australia and most international markets, excluding restricted jurisdictions under…
148
149
When static makes your C code 10 times faster
Addendum : Many Hacker News commenters correctly point out that the right way to indicate to the compiler that a value does not change is const . This post is not about C best practices, but just about how one can easily get to the bottom of this kind of surprising behavior by looking at the compiler output. This…
150