Hello again.

Comic of the week

Links of the week

Things I Wished More Developers Knew About Databases

In data-heavy systems, databases are at the core of system design goals and tradeoffs. Even though it is impossible to ignore how databases work, the problems that application developers foresee and experience will often be just the tip of the iceberg. In this series, I’m sharing a few insights I specifically found useful for developers who are not specialized in this domain.

How to Effectively Manage Low Performers: The CARES Framework

As a manager, one of the most challenging tasks is to help low-performing team members improve their motivation and skill levels. It’s essential to understand that these team members are still valuable and can contribute significantly to the team’s success with the right guidance. To assist managers in this endeavor, I introduce the CARES framework — a five-step process designed to transform low performers into key contributors.

Mitigating Attack Vectors in GitHub Workflows

GitHub Actions are commonly used to automate processes in repositories, by running CI (continuous integration) tests on pull requests for example. It can also be used to make a package release process more secure just by making it automated. But, it is important to be careful to ensure that they are safe and do not expose the project to attacks. Understanding how workflows can be part of the attack surface for a GitHub project helps us understand how to prevent it.

Do you want to be the first to know about new posts? Subscribe now for FREE. Just cool content, no SPAM!

Subscribe now

Understanding pgvector's HNSW Index Storage in Postgres

Creating a vector index with pgvector is straightforward - just run CREATE INDEX ON t USING hnsw(col vector_l2_ops). But what is actually going on under the hood as we run this and insert or modify data?

In this article, we'll take a deep dive to understand the underlying index file created by pgvector in Postgres.

Continuous reinvention: A brief history of block storage at AWS

We started with software virtualization. Until late 2017 all EC2 instances ran on the Xen hypervisor. With devices in Xen, there is a ring queue setup that allows guest instances, or domains, to share information with a privileged driver domain (dom0) for the purposes of IO and other emulated devices. The EBS client ran in dom0 as a kernel block device. If we follow an IO request from the instance, just to get off of the EC2 host there are many queues: the instance block device queue, the Xen ring, the dom0 kernel block device queue, and the EBS client network queue. In most systems, performance issues are compounding, and it’s helpful to focus on components in isolation.

"SRE" doesn't seem to mean anything useful any more

To me, a SRE is *both* a sysadmin AND a programmer, developer, whatever you want to call it. It's a logical-and, not an XOR.

By sysadmin, I mean "runs a mean Unix box, including fixing things and diving deeply when they break", and by the programmer/whatever part of it, I mean "makes stuff come into existence that wasn't there before". In particular, I expect someone to run the *right* things on those boxes, to find the actual problems and not just reboot stuff that looks squirrelly, and that they write good, solid code that's respectful of the systems and the network. They probably write programs to make the sysadmin part of the job run itself. Automation for the win.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

Dependency Injection Principles, Practices, and Patterns

Subscribe now

Have a nice week. 😉

Have you read last week's post? Check the archive.

Hello again.

Links of the week

story points are pointless, measure queues

The problems come when people hear, "This will take 2 weeks." Inevitably, this creates confusion because when a dev team estimates the time they often mean "This will take 2 weeks with our entire team if we have no interruptions once we start on the work next month after this other dependency has been resolved..." which isn't written into the criteria. Other work is going to come up on a live system. When you say 2 weeks to people on the business side of the house, they often hear "2 weeks" without any of the other ceremony.

Hiding in plain sight: Modifying process names in UNIX-like systems

Ok, time to get back to the land of Linux. Let's take a look now "under the hood". The intention of here is to understand the internal data structures at play. Refer to the following diagram which depicts various internal structures within the Linux kernel when a python script was run from the command line.

Two Threads, One Core: How Simultaneous Multithreading Works Under the Hood

Simultaneous multithreading (SMT) is a feature that lets a processor handle instructions from two different threads at the same time. But have you ever wondered how this actually works? How does the processor keep track of two threads and manage its resources between them?

Do you want to be the first to know about new posts? Subscribe now for FREE. Just cool content, no SPAM!

Subscribe now

LISP's Grandfather Paradox

John McCarthy based Lisp on general recursion, but devised it to operate on symbols rather than numbers. He was interested in AI, where concepts, the way humans manipulate them, occupy the center stage. A symbol can stand in for any idea, whether code or data. Just today we've seen just how expressive symbolic expressions are, effortlessly encoding the formulas of primitive recursion. Somehow we have made Lisp revisit its conceptual ancestor. We've traveled in time like in the grandfather paradox. Not with the intent to kill the progenitor, but rather to embrace him, bring him back to life, even if for a moment.

Modeling B-trees in TLA+

I was reading along with the book as Petrov explained how B-trees work, and nodding my head. But there’s a difference between feeling like you understand something (what the philosopher C. Thi Nguyen calls clarity) and actually understanding it. So I decided to model B-trees using TLA+ as an exercise in understanding it better.

Parsing Protobuf Definitions with Tree-sitter

If you work with Protocol Buffers (protobuf), you can really save time, boredom, and headache by parsing your definitions to build tools and generate code. […] In the process I realized how awesome the Tree-sitter parsing library is and that it probably was going to support extracting everything I wanted to get from our protobuf definitions. Neovim uses Tree-sitter extensively.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

How Big Things Get Done: The Surprising Factors That Determine the Fate of Every Project, from Home Renovations to Space Exploration and Everything In Between

Subscribe now

Have a nice week. 😉

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

No sleepwalk to success

I’ve spent 20 years investing in and working with deeply technical founders at a very early stage. Looking back, there seems to be one overwhelming determinant of success for those technical teams. In my experience, success is highly correlated with how obsessive a founding team is about developing an extraordinarily efficient go-to-market (GTM) machine.

Testability = Modularity

Testability is the quality which enables the developer to easily test a software component separately from the rest of the architecture.

Thus, a testable architecture is one which is made of components that are easily testable in isolation.

How do the latter sentences relate to modularity?

Home-Cooked Software and Barefoot Developers

For the last ~year I've been keeping a close eye on how language models capabilities meaningfully change the speed, ease, and accessibility of software development. The slightly bold theory I put forward in this talk is that we're on a verge of a golden age of local, home-cooked software and a new kind of developer – what I've called the barefoot developer.

Do you want to be the first to know about new posts? Subscribe now for FREE. Just cool content, no SPAM!

Subscribe now

The sad state of property-based testing libraries

The first project at Ericsson that Quviq helped out testing was written in Erlang. Unlike Haskell, Erlang is not a pure functional programming language and on top of that there’s concurrency everywhere. So even the second, monadic, version of QuickCheck didn’t turn out to be ergonomic enough for the job. This is what motivated the closed source Quviq QuickCheck version written in Erlang, first mentioned in the paper Testing telecoms software with Quviq QuickCheck (2006). The main features of the closed source version that, as we shall see, are still not found in many open source versions are:

1. Sequential stateful property-based testing using a state machine model;

2. Parallel testing with race condition detection by reusing the sequential state machine model.

Simple event broker tries Tiger Style

It has taken me much longer to learn this than is reasonable, but I now finally know, and act as if I know!, that the very first thing you must do when trying to make your program faster is to measure it and be very systematic about your measurements. Yes, I know it is much more fun to guess at the problem and try out random solutions, crossing your fingers in the hope that one of your guesses magically make things go brrr. But if you plan to make progress instead of trying your luck all day, going straight to some sort of profiling is the winning move. Every. Single. Time. Even if you’re just printf’ing timestamps; you must measure!

gRPC Over HTTP/3

In this post, we’ll dive into what HTTP/3 is and explore the compelling reasons why it’s an ideal fit for gRPC applications. We’ll uncover the technical advancements that promise to make HTTP/3 faster, more reliable, and more secure. But we won’t stop at theory; we’ll get our hands dirty with practical examples in Go, demonstrating how to set up and test gRPC servers and clients over HTTP/3.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

Joshua Weissman: Texture Over Taste

Subscribe now

Have a nice week. 😉

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

The Death of the Junior Developer

I wrote this post a week ago and have been thinking hard about whether I believe the premise, which is that within a few years, the norm for source code will be that it is written and modified by LLMs via prompting. For all practical purposes, all source code will be written this way, with exceptions becoming ever rarer.

Not only do I believe it, I could even see it happening in 12-18 months at the current rate of LLM progress. I think the change will have a ton of fallout, only some of it foreseeable. And one casualty might well be junior devs, in the sense that they become less marketable and it could cause various kinds of crunches across the industry.

A write-ahead log is not a universal part of durability

A database does not need a write-ahead log (WAL) to achieve durability. A database can write its long-term data structure durably to disk before returning to a client. Granted, this is a bad idea! And granted, a WAL is critical for durability by design in most databases. But I think it's helpful to understand WALs by understanding what you could do without them.

Why do we underestimate how long it will take to complete a task?

The planning fallacy stems from our overall bias towards optimism, especially when our own abilities are concerned. In general, we are oriented towards positivity. We have optimistic expectations of the world and other people; we are more likely to remember positive events than negative ones; and, most relevantly, we tend to favor positive information in our decision-making processes.

Do you want to be the first to know about new posts? Subscribe now for FREE. Just cool content, no SPAM!

Subscribe now

Gratification

The only goal I have with this article is to make you think about the consequences of new and innovative tools and frameworks. Particularly if they are immediately compelling, they may be empty calories. Consider if there may be disadvantages to adopting a new way of doing things.

Integrate Go Library into a JavaScript Webpage with WebAssembly

With WebAssembly you can integrate different programming languages together. This opens up new possibilities to use great libraries written in one language in another different platform. In this article I want to show how to integrate a Go library for sentence tokenization in a JavaScript webpage. Normally, you cannot run Go code in a web browser, but with the WebAssembly technology you can.

A Git story: Not so fun this time

Monotone was created by Graydon Hoare. In 2001, Graydon, who lived in Canada, wanted to work more easily with his Australian friend. So they developed a system similar to today’s Continuous Integration (CI), which wasn’t widely known yet. Their system ensured that the code always passed tests.

In 2002, Graydon became interested in combining version control with CI. At that time, only Aegis had such a concept. Graydon also saw his friends using BitKeeper and thought that merging Aegis with DVCS could be an opportunity. That’s how Monotone came into being.

Remarkably, Graydon later joined Mozilla and created the Rust programming language.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

Mind Over Muscle: Writings from the Founder of Judo

Subscribe now

Have a nice week. 😉

Have you read last week's post? Check the archive.

It’s been about 2 years since I started the Polymath Engineer Weekly newsletter. We’re finally at issue #100! 🥳🍻🎉

You have no idea how many things I had to read that don’t even get to show up here. I try to bring mostly high quality information, filtering out a lot of garbage out there… I’m glad I can consistently read cool stuff and reach you all in more than 20 countries. I’m certainly becoming a better engineer through it.

Thanks for your attention and all the feedback, I hope you are enjoying the ride.

No recap of previous posts or shit like that. Let’s get to this week’s content!

Comic of the week

Links of the week

The Programmer's Brain

Learn how to optimize your brain’s natural cognitive processes to read code more easily, write code faster, and pick up new languages in much less time. This book will help you through the confusion you feel when faced with strange and complex code, and explain a codebase in ways that can make a new team member productive in days!

Election Monoids And “Equal” Votes

I care a lot about the best ways to run elections. I also care about mathematics, and algebra in particular. What happens when you mix the two? Let’s find out!

Behind AWS S3’s Massive Scale

At the core of all of this are the storage node servers. They are simple key-value stores that persist object data to hard disks. The node servers only shards of the overall object data, with the control plane replicating said shards across many different nodes.

AWS has written a lot about their newly-written storage backend - ShardStore. It initially started with just 40k lines of Rust.

Under the hood, it’s a simple log-structured merge tree (LSM Tree) with shard data stored outside the tree to reduce write amplification.

Do you want to be the first to know about new posts? Subscribe now for FREE. Just cool content, no SPAM!

Subscribe now

DDD re-distilled

This article is an abstract of the book DDD Distilled written by Vaughn Vernon

How does a computer/calculator compute logarithms?

As I was learning how to script in Python, I began to get curious about how things worked under the hood of this programming language. Since Python is open source, I was able to look at its source code to see how different mathematical functions, particularly in its math and cmath modules, worked and how their code was structured. (…)
However, when it came to the basic complicated math functions, like the logarithm, trigonometric, and exponential functions, I found that the Python source code didn’t have them written up, because Python used the functions from the C library. This led me to a new adventure where I looked into the original C code, which, while more difficult to read, is surprisingly readable once you have learned Python.

What the F*%k is a moment arm and what does it have to do with my training!?

Notice how the person with long legs and a short torso has a longer moment arm between the bar and the knee and how the knees go past the toes? Imagine, if we were to give this lifter a more upright torso – the knees would be much farther past the toes in order for them to get their hips down further while driving their chest up

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

The Little Typer

Subscribe now

Have a nice week. 😉

Have you read last week's post? Check the archive.

Hello again. Almost 100 posts, let’s go!

Comic of the week

Links of the week

Secure Enclave

The Secure Enclave is a dedicated secure subsystem integrated into Apple systems on chip (SoCs). The Secure Enclave is isolated from the main processor to provide an extra layer of security and is designed to keep sensitive user data secure even when the Application Processor kernel becomes compromised. It follows the same design principles as the SoC does—a boot ROM to establish a hardware root of trust, an AES engine for efficient and secure cryptographic operations, and protected memory.

I’ve Stopped Using Box Plots. Should You?

I used to use box plots often, but I rarely use them now. In fact, virtually never. Why not? Well, if you watched the explanatory video linked above, you saw that it took me over four minutes to explain how to read a box plot. If a chart type requires that much time to explain, it had better deliver a huge “epiphany payoff,” as data storytelling expert Brent Dykes calls it, i.e., it had better deliver valuable insights that can’t be communicated using simpler, more familiar chart types that require less time and effort for the audience to understand.

How to use Prometheus for anomaly detection in GitLab

One of the more basic functions of the Prometheus query language is real-time aggregation of time series data. Andrew Newdigate, a distinguished engineer on the GitLab infrastructure team, hypothesized that Prometheus query language can also be used to detect anomalies in time series data.

Andrew broke down the different ways Prometheus can be used for the attendees of Monitorama 2019. This blog post explains how anomaly detection works with Prometheus and includes the code snippets you’ll need to try it out for yourself on your own system.

Do you want to be the first to know about new posts? Subscribe now for FREE. Just cool content, no SPAM!

Subscribe now

How to Build a Cybersecurity Career

Having a lab is essential. It’s actually one of the first things I ask when I’m looking at candidates during interviews. I ask what kind of lab or network they have to play with, and if they reply that they don’t have either I thank them for their time.

The lab is where you learn. The lab is where you run your projects. The lab is where you grow.

the Gilbert–Johnson–Keerthi algorithm explained as simply as possible

The GJK algorithm is a weird way to do a simple thing.

We have shape A and shape B, and we'd like to determine if they overlap.

A shape is a set of infinitely many points. If there exists any point that's a member of both sets, then the shapes overlap.

What are Genetics for working out and bodybuilding explained with Hypertrophy Coach Joe Bennett

This video explores the role of genetics in bodybuilding. He explains what genetics are and how they affect muscle growth potential, body structure, and even muscle insertions. He also discusses how to overcome genetic limitations through hard work and consistency.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

A Philosophy of Software Design

Subscribe now

Have a nice week. 😉

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

Deep Dive into Anthropic’s Sparse Autoencoders by Hand ✍️

Sparse Autoencoders help break down the problem of ‘polysemanticity’ — neural activations that correspond to several meanings/interpretations at once by focusing on sparsely activating features that hold a single interpretation — in other words, are more one-directional.

DuckDB Doesn’t Need Data To Be a Database

Databases have gotten so good at this, that the term is almost misleading now. “Base” suggests something rigid, without which the data would slip away. But the data is always there, just bits on a nameless hard disk. The structure and the accessibility that a modern database provides exist completely independently from that hard disk. That’s right – most databases no longer have any data in them.

CPython Garbage Collection: The Internal Mechanics and Algorithms

The Python code executes on the interpreter in the context of an OS thread, and the runtime maintains the information about the thread’s state in an object called thread state. The thread state contains information about the code executing on the interpreter in the context of that thread, such as the stackframe pointer, the GC state, and the eval_breaker.

Do you want to be the first to know about new posts? Subscribe now for FREE. Just cool content, no SPAM!

Subscribe now

The TTY demystified

The TTY subsystem is central to the design of Linux, and UNIX in general. Unfortunately, its importance is often overlooked, and it is difficult to find good introductory articles about it. I believe that a basic understanding of TTYs in Linux is essential for the developer and the advanced user.

Go evolves in the wrong direction

Since Go1.23, the for ... range loops can be applied to functions with special signatures (aka pull and push functions). This makes impossible to understand what the given innocent for ... range loop can do under the hoods by just reading the code. It can do anything, like any function call can make. The only difference that the function calls in Go were always explicit, e.g. f(args), while for ... range loop hides the actual function call. Additionally, it applies non-obvious transformations for the loop body

How We Made PostgreSQL as Fast as Pinecone for Vector Data

The DiskANN algorithm was developed by work coming out of Microsoft. Its goal was to store a very large number of vectors (think Microsoft scale). At that scale, it was simply uneconomical to store everything in RAM. Thus, the algorithm is geared towards enabling storing vectors on SSDs and using less RAM. Its details are described very well in the paper, so I’ll only give a bit of intuition below.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

Skin in the Game: Hidden Asymmetries in Daily Life

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

Service level adoption obstacles

I have tried Datadog, Grafana, Elastic, and a few others. NO ONE GOT IT RIGHT! It is as if they just wanted to check a box to add SLO as a feature to their products. SLO is not just a feature. It is THE way you add value to empower reliability engineering, reshape the organization, and build a culture of data-driven optimization.

Controlling the Taylor Swift Eras Tour wristbands with Flipper Zero

Many large concerts feature wristbands that light up on command. They are used to produce varied visual effects across a stadium. One company that makes these is PixMob. Their controllable, illuminated wristbands are currently being used as part of Taylor Swift's Eras Tour.

2D Rigid Body Collision Resolution

From Mario bouncing off a Goomba to two cars bumping into each other in a racing game, dealing with collisions is such an integral part of most video games that we often take it for granted.

Pretty cool post. Lots of interactive examples. Give it a try!

Subscribe now

6 Archetypes of Broken Ownership

The rationale is:

1. You cannot be responsible for something you don’t control. You need the mandate.

2. You cannot use that mandate effectively over something you don’t understand. You need knowledge.

3. You gain knowledge only if you are fully responsible for the consequences of your mandate.

learning a new language is easy, actually

While some languages take as little as 600 hours to reach fluency, you can also just watch videos about how to learn the language for 1000 hours. Should achieve the same effect, honestly.

The $400 BILLION Dollar Scam DESTROYING Your Feet

In this video I’ll give you the 3 Step System for Strong Feet. This is extremely important because unfortunately “BIG SHOE” AKA Nike, Vans, and Adidas have been destroying our feet for decades. Without us realizing it, we’ve continued to give our dollars to these shoe Juggernauts, who sell us a product that puts us in pain, leading us to movement dysfunction.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

The Flavor Bible: The Essential Guide to Culinary Creativity, Based on the Wisdom of America's Most Imaginative Chefs

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

Fundamentals

A common lament among software developers is that the pace of technology is so overwhelming that they can't keep up. This is true. You can't keep up.

There will always be something that you don't know. In fact, most things you don't know. This isn't a condition isolated only to technology. The sum total of all human knowledge is so vast that you can't know it all. What you will learn, even after a lifetime of diligent study, will be a nanoscopic fraction of all human knowledge - even of everything related to software development. You can't stay current. Get used to it.

A more appropriate question is: How do I keep my skill set relevant?

The Art of System Debugging — Decoding CPU Utilization

It is the API Gateway for Flipkart which calls a myriad of microservices using the scatter gather pattern. It uses circuit breaking to avoid thundering herd issues on upstream dependencies. All frontend clients like the Desktop, Mobile Site, Android and iOS apps consume this API to render the interfaces. For the sake of this blog post, let’s call this application the APIGateway application.

Unemployed Guy Explains Option Trading in 5 Levels of Complexity

My name is Ben, I'm an unemployed option trader and youtube finance professional with 0 certification, training, or formal education.

Subscribe now

The Insane Engineering of the Gameboy

It’s awesome to see how hard constraints can foster very creative solutions.

Mapping the Human Brain and Where This May Lead Us

The idea is that in order to understand the brain, we must understand the network and not merely the individual neurons or functioning of individual synapses. This is similar to how our understanding of a genome relies on observing how DNA is transcribed into functional building blocks; components which ultimately result in anything from as simple as a single-celled eukaryote, to a functioning human body, or any other lifeform. By observing the way the genome is used in this context, we gain understanding of how both the genome and the cell or body it’s part of fit together.

Test clocks: How we made it easier to test Stripe Billing integrations

Billing integrations are also difficult to test. Historically, the only way to run a test was to wait for time to pass—typically by creating test configurations with shorter subscription cycles than real production systems, or running 10-second trials to force subscriptions to cycle—and then look for any bugs that might surface in the course of normal business. Of course, doing anything that is not a perfect mirror of your production system is a shaky foundation to build on.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

How to Know a Person: The Art of Seeing Others Deeply and Being Deeply Seen

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again. We have some pretty cool stuff this week 😎

Comic of the week

Links of the week

The Internet is starting to Break - Here's Why.

Why big tech companies suck right now. Looking at you Amazon, Netflix, Uber, Spotify, Facebook, Snapchat, Instagram, TikTok, YouTube...

Datomic Pro 1.0.7075

Both log and indices are stored as persistent, immutable trees in a data store like Cassandra or DynamoDB. Because tree nodes are immutable, their backing storage only needs to guarantee eventual consistency. A small pointer to the roots of these trees provides a consistent, immutable snapshot of the database’s state. To commit a transaction, a transactor saves new immutable tree nodes to storage, then executes a compare-and-set (CaS) operation to advance the root pointer. This CaS operation must execute under Sequential consistency.

Even if you are not interested in Datomic or Clojure, give this article a shot. There is a lot of Distributed Systems knowledge distilled into it. I’m sure you won’t regret it.

It’s Lambdas All the Way Down

The fact that you can start with a few building blocks and make a whole programming language that allows you to run any computation there ever will be is in itself mind blowing. But did McCarthy design those primitives, or as PG puts it, he discovered them? As it turns out, those building blocks could themselves be broken down into smaller pieces. Think of it like atoms in the universe that can be broken into sub-atomic particles. Surprisingly, these particles were found decades ago by Alonzo Church, the inventor of Lambda calculus long before electronic computers even existed!!

Subscribe now

The Algorithm Behind Jim Simons's Success

Simons lived life on his own terms. The math prodigy became a code breaker for the NSA before chairing the Math Department at Stony Brook. He then abandoned his career in academia to build a hedge fund. Simons’s life unfolded in an unpredictable yet remarkable way because he followed his curiosity to interesting problems.

The Mysterious AI Reading List: Ilya Sutskever's Recommendations

John Carmack, the renowned game developer, rocket engineer, and VR visionary, shared in an interview that he asked Ilya Sutskever, OpenAI co-founder and former Chief Scientist, for a reading list about AI. Ilya responded with a list of approximately 40 research papers, saying:

If you really learn all of these, you’ll know 90% of what matters today.

Finality does not exist in payments

Younger generations might not appreciate this, but there was a very real question in the late 1990s (and across much of the world today) whether transactions could ever take place over the Internet without being able to look counterparties in the eye. “Aren’t hackers just going to take all money exposed to the Internet?” And low and behold, they did not, but a necessary precondition for answering that question was there being money exposed to the Internet. Credit card guarantees of reversibility substantially made that happen.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the week

Measure What Matters: How Google, Bono, and the Gates Foundation Rock the World with OKRs

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

Deno KV internals: building a database for the modern web

We chose to build Deno KV atop of FoundationDB, Apple’s open source distributed database used in iCloud and by Snowflake, since it’s perfect for building scalable database solutions: it’s thoroughly verified with deterministic simulation, scalable and efficient, and provides the transactional key-value storage API.

Maker vs. Manager: How Your Schedule Can Make or Break You

The numerous articles about the routines of successful people miss the point. Waking early doesn’t make you a novelist. Segmenting your day doesn’t make you an entrepreneur.

The lesson isn’t what time to set your alarm, which journalling prompt to use, or how long to wait before drinking your coffee. The lesson is that different work requires different schedules.

Chapter I: The birth of Parquet

During that time, the data platform team at Twitter had a paper reading group. People would read papers and present them to the group if they were interesting. At the time I re-read the Dremel paper and presented to the group. I also read interesting papers that described how those MPP databases work: C-Store, Vertica, Monet-DB and others. This led to a better understanding of distributed query engines and the benefits of column stores.

Subscribe now

Consistency Large Language Models: A Family of Efficient Parallel Decoders

LLMs have been traditionally regarded as sequential decoders, decoding one token after another. In this blog, we show pretrained LLMs can be easily taught to operate as efficient parallel decoders. We introduce Consistency Large Language Models (CLLMs), a new family of parallel decoders capable of reducing inference latency by efficiently decoding an n-token sequence per inference step.

The Rise of the Forever Renter Class

I am a part of what you might call the Forever Renter class. This is a cohort of people in their mid-20s to late-30s who are either unwilling or unable to buy a home under current market conditions. Unfortunately, those conditions aren’t moving in the right direction either.

Tracking Illicit Brazilian Beef from the Amazon to Your Burger

With grassroots support from labor unions and Indigenous communities, he had mapped the complex networks of cattle farms responsible for illegal deforestation. He then tracked the often-illicit beef through JBS’s slaughterhouses and packing plants to the freezers, shelves, and customer trays of retail outlets and fast-food restaurants around the world. When his sleuths were done, the fingerprints of forest destruction were plain to see. Six of Europe’s biggest retail chains reacted by halting purchases of JBS beef.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the Week

The Art of Cyberwarfare: An Investigator's Guide to Espionage, Ransomware, and Organized Cybercrime

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

10 regrets of experienced programmers

I recently asked "You are on your deathbed, what do you regret the most about programming?". Let's break down the biggest mistakes shared by most software engineers.

Neon - Serverless PostgreSQL

Neon is composed of a stateless compute layer and a disaggregated storage layer. Postgres instances form the stateless compute layer, and the storage layer is itself disaggregated and consists of a fleet of storage nodes where each node stores the data of multiple tenants. The fleet is broken into horizontally scalable services for the write-ahead logs (WAL service) and data pages (Page service).

Multicylinder Engine Simulation

In this video, I continue working on Engine Simulator, pushing the fluid simulation to its performance limits with 4 cylinder engines.

Subscribe now

Graceful shutdown in Kubernetes

If the nature of those pods is so ephemeral, what happens when a pod is in the middle of responding to a request but is told to shut down?

Is the request completed before shutdown?

What about subsequent requests? Are those redirected somewhere else?

Advanced Retriever Techniques to Improve Your RAGs

We know that when we apply a “Naive retrieval”, we are calculating the similarity of all the chunks of the vector database with the query. The more chunks the vector store has, the more similarity calculations will have to be done. Now, imagine being able to do a prior filter based on the metadata, and after selecting the chunks that meet the conditions imposed in relation to the metadata, calculate similarities. This can drastically reduce computational and time cost.

Proof Resting Doesn't Keep Meat Juicy

Everyone knows you have to rest your steaks. I used to believe that, too. That’s the thing about science: it doesn’t matter what you believe, if you experiment diligently and follow the data, the truth will come out. Watch this video and you'll also learn how to manage carryover proactively. You’ll never have to suffer through another sad, overcooked steak for fear of losing the juices.

I’m not fully convinced yet. I guess I should run my own experiments and see it for myself. Although it seems that water content is similar between both, in my experience I find a difference in final texture and perception of “juiciness”.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the Week

Practical Social Engineering: A Primer for the Ethical Hacker

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

‘Lavender’: The AI machine directing Israel’s bombing spree in Gaza

The Israeli army has marked tens of thousands of Gazans as suspects for assassination, using an AI targeting system with little human oversight and a permissive policy for casualties.

“At 5 a.m., [the air force] would come and bomb all the houses that we had marked,” B. said. “We took out thousands of people. We didn’t go through them one by one — we put everything into automated systems, and as soon as one of [the marked individuals] was at home, he immediately became a target. We bombed him and his house.”

Kora - Serverless Kafka

The architecture of Kora is one of a light proxy layer over a fast fault-tolerant cache that offloads data asynchronously to cloud object storage as the primary data store. As I described in the series introduction, cloud object storage presents both a large opportunity for reducing storage costs but also comes with the significant downside of high request latency and an economic model that punishes small requests

Quant trading's most lucrative programming language (you never heard of it)

In today's video I interview Jeremy Lucid. Jeremy programs in one of the most niche and lucrative programming languages in the world of quantitative trading. While most have never heard of it, all large banks and cutting-edge financial institutions are demanding ever larger amounts of developers proficient in this language (despite their continued shortage).

Subscribe now

The columnar roadmap: Apache Parquet and Apache Arrow

Vertical integration from storage to execution greatly improves the latency of accessing data by pushing projections and filters to the storage layer, reducing time spent in IO reading from disk, as well as CPU time spent decompressing and decoding. Standards like Arrow and Parquet make this integration even more valuable as data can now cross system boundaries without incurring costly translation. Cross-system programming using languages such as Spark, Python, or SQL can becomes as fast as native internal performance.

Comparing option pricing methods in q

Within the financial industry, there is a need to price complex financial instruments. Despite this need, there are a lack of analytical solutions to do so. MC methods are used with the financial industry to mimic the uncertainty associated with the underlying price of an instrument and to subsequently generate a value based on the possible underlying input values. One example of where MC is used in finance, is in evaluating an option on an equity.

Talking to memory: Inside the Intel 8088 processor's bus interface state machine

The bus state machine is an example of how the 8088's design consists of complications on top of complications. While the four-state bus cycle seems straightforward at first, it gets more complicated due to prefetching, wait states, the HALT instruction, and the bus hold feature, not to mention the interactions between these features. While there were good motivations behind these features, they made the processor considerably more complicated. Looking at the internals of the 8088 gives me a better understanding of why simple RISC processors became popular.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the Week

The Oxygen Advantage: Simple, Scientifically Proven Breathing Techniques to Help You Become Healthier, Slimmer, Faster, and Fitter

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again. I think this week's links are pretty cool, enjoy ;)

Comic of the week

Links of the week

Paper Notes: f4 – Facebook’s Warm Blob Storage System

Blob storage system is built in such a manner that it abstracts away the concept of hot/warm storage from the client. Requests enter the system through a router layer where create requests are sent to the hot layer whereas delete requests are sent to the host where the blob is currently being stored. For read requests, router extracts the logical volume id & map it to the physical volume where blob is stored.

Arena allocator tips and tricks

An arena is a memory buffer and an offset into that buffer, initially zero. To allocate an object, grab a pointer at the offset, advance the offset by the size of the object, and return the pointer. There’s a little more to it, such as ensuring alignment and availability. We’ll get to that. Objects are not freed individually. Instead, groups of allocations are freed at once by restoring the offset to an earlier value. Without individual lifetimes, you don’t need to write destructors, nor do your programs need to walk data structures at run time to take them apart. You also no longer need to worry about memory leaks.

Formal Methods: Just Good Engineering Practice?

At first, this may seem counter-intuitive. Formal methods aren’t cheap, aren’t particularly easy, and don’t fit well into every software engineering approach. Its reasonable to start with the belief that a formal approach would increase costs, especially non-recurring engineering costs. My experience is that this isn’t true, for two reasons. The first is rework. Software engineering is somewhat unique in the engineering fields in that design and construction tend to happen at the same time, and a lot of construction can be started without a advancing much into design.

Subscribe now

Tableflow: The Stream/Table, Kafka/Iceberg Duality

The data infrastructure world is undergoing multiple paradigm shifts simultaneously, all driven by S3 and the other cloud object storage services (GCS, Azure Blob Storage). While the trends we are seeing now have been in the making for a number of years, it seems that 2024 will be the year that the shift really gathers pace. As they say, change comes slowly, then all at once.

How do One-Time Passwords Work?

The algorithm itself does the following three steps. It generates a HMAC-SHA-1 value, where secret is the key and counter is the message to be hashed using the key. Then it truncates the 20-byte hash string to a shorter value.

The truncation is rather uncomplicated, the last byte from the 20-byte string is taken, the lower 4 bits are considered the “offset”. This makes the offset to be always in the range from 0 to 15.

Inside the Super Nintendo Cartridges

One of the remarkable characteristics of the Super Nintendo was the ability for game cartridges (cart) to pack more than instructions and assets into ROM chips. If we open and look at the PCBs, we can find inside things like the CIC copy protection chip, SRAM, and even "enhancement processors".

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the Week

Rust for Rustaceans: Idiomatic Programming for Experienced Developers

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

Why is observability so expensive?

Observability vendors will gladly identify any number of other causes such as “It’s important and should be 30% of one’s budget!” and “Stop using metrics and just use events!”, but the fundamental reality as was said previously is that there is simply no free lunch: massive production of telemetry at the source, without regard for the usefulness of the data, will inevitably lead to high costs and unhappy end-users. Someone has to pay for all of that transport and storage.

The Real Threat To Your Job Now Is Not AI

Picture this: you're diligently working on a project, navigating through lines of code like a seasoned explorer. Suddenly, a colleague swoops in, effortlessly solves a complex problem in minutes that had you scratching your head for hours, and then proceeds to refactor the entire module with surgical precision. You've just encountered a 10x programmer in action.

Microservices Design Principles You Really Need To Learn

As you apply these principles in your own projects, don't be afraid to experiment and iterate. Each application is unique, and what works for one may not work for another. By embracing a mindset of continuous improvement and learning from both successes and failures, you'll be well on your way to mastering the art of microservices design.

Subscribe now

Multithreading and the Memory Subsystem

In this post we investigate how the memory subsystem behaves in an environment where several threads compete for memory subsystem resources. We also investigate techniques to improve the performance of multithreaded programs – programs that split the workload onto several CPU cores so that they finish faster.

Event-Driven Architecture Fundamentals and Common Pitfalls (and How to Avoid Them)

Designing event payloads with all of the data at the time of the event is often considered a prudent thing to do. However, if the data changes between the time that the event message is published and the message is processed by the receiver, the processing will be based upon stale data. Instead, design the event message with limited data to prevent sharing stale data. Consider details such as the field that was changed, the old and new value, and any additional details that the message receiver will use to determine if further action needs to be taken or if the event should be ignored.

Product-Market Fit Isn’t a Black Box — A New Framework to Help B2B Founders Find It, Faster

“From the outside, you can romanticize a founder’s life and think, ‘The world's your oyster. You can build whatever you want. You're going to start a company, it'll be great. You get to decide everything.’ And that's all true, but in the day-to-day, it's more like, ‘What am I doing? Does anyone care? Will anyone ever care?’ There’s so much in the early days where you can't tell if you’re doing it wrong. Having a group of other people going through a similar experience to talk about that with and normalize some of the uncertainty is really, really helpful. It can help you figure out if it’s an idea that you’re pulled toward because you’re just looking for something in all of the uncertainty, or if it’s actually viable.”

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the Week

Hacking APIs: Breaking Web Application Programming Interfaces

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

Garbage Collection for Systems Programmers

It’s just not true. It’s ideology. And I bought it for over a decade until I joined a team that builds systems—systems people bet their lives on—that provide sub-microsecond latency, using a garbage-collected language that allocates on nearly every line. It turns out modern GCs provide amazing throughput, and you don’t need to throw that out for manual memory management just because some of your system absolutely needs to run in n clock cycles.

Containers and Unikernels: Conceptually Similar, Fundamentally Different, and Inextricably Intertwined

Whilst unikernels were born out of the security domain and containers out of reproducibility (achieved through isolation), today, containers and unikernels still share a core principle: software encapsulation. Both create isolated environments for applications. Containers achieve this by sharing the host kernel but isolating the application’s processes, libraries, and files. Unikernels, on the other hand, go a step further. They bundle the application code with a minimal pre-configured kernel, eliminating the need for a full host kernel altogether.

Improving how credit cards work under the covers

Remember, we don’t care about PANs because they are account numbers. We care about them because they are presumptive evidence of an agreement by a customer and a business to establish a billing relationship. A token is much stronger evidence, because a) they can only be created by a real business actually operating the network machinery to talk to an issuer and b) they can be repudiated at will by a business, issuer, or (de facto) by a consumer, without consequence to other uses of the underlying account. You may have had the experience of having a card stolen and needing to get the bank to give you a new account number, and then spending hours informing companies about the new number. If a token gets repudiated, it is non-event for the customer, the bank, and every other organization in the world.

Subscribe now

How Tech Giants Cut Corners to Harvest Data for A.I.

OpenAI was desperate for more data to develop its next-generation A.I. model, GPT-4. So employees discussed transcribing podcasts, audiobooks and YouTube videos, the people said. They talked about creating data from scratch with A.I. systems. They also considered buying start-ups that had collected large amounts of digital data.

OpenAI eventually made Whisper, the speech recognition tool, to transcribe YouTube videos and podcasts, six people said. But YouTube prohibits people from not only using its videos for “independent” applications, but also accessing its videos by “any automated means (such as robots, botnets or scrapers).”

OpenAI employees knew they were wading into a legal gray area, the people said, but believed that training A.I. with the videos was fair use. Mr. Brockman, OpenAI’s president, was listed in a research paper as a creator of Whisper. He personally helped gather YouTube videos and fed them into the technology, two people said.

The Design Philosophy of Great Tables

Tables emerged from square grids. When grids are made like this, you invariably generate containers that may hold some sort of information. The earliest known examples of grids go very far back in human history. Twenty-five thousand year old representations of the grid are found on the walls of the Lascaux and Niaux caves in France1.

Why Exhausts Make So Much Power😵

As critical as this Exhaust Systems are, its the most overlooked, commonly just pieced together with no method of operation and it can be the difference between power being lost or gained depending on the application. In this day and age you have the ability to have it all, all the sound, all the performance and all the comfortability configured into one system and In this video we will reveal the truths and falsehoods, myths and methodologies behind building the perfect exhaust for your vehicle

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the Week

Forever Strong: A New, Science-Based Strategy for Aging Well

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again. Hey, please answer the survey down below. It’ll help me a lot to bring relevant content to you ;)

Comic of the week

Links of the week

Understanding Jane Street

Jane Street is a quantitative prop-trading firm, i.e. they use their internal capital to make markets in liquid assets around the world, balancing completely automated strategies with human intervention. In terms of expected value, this might be the single best field for a smart and technical person to work in. […]

The history of finance in general and quantitative finance in particular is littered with examples of very smart people who had good track records and then blew up. The brilliant people are table stakes. Building a lasting institution means embedding risk awareness into everything.

⏯ But what is a GPT? Visual intro to Transformers

I’m a simple man. 3Blue1Brown posts a video, I watch it.

It's hearty, it's meaty, it's mold - Hacking the genome of fungi for smart foods of the future

Hill-Maini's next objective is to make the fungi even more appealing by tuning the genes that control the mold's texture. "We think that there's a lot of room to explore texture by varying the fiber-like morphology of the cells. So, we might be able to program the structure of the lot fibers to be longer which would give a more meat-like experience. And then we can think about boosting lipid composition for mouth feel and further nutrition"

Subscribe now

⏯ Go: Pointers for Performance?

You should not use pointers solely on the assumption they will improve performance. When using runtime languages, sometimes the trade-offs are not obvious or intuitive.

Copying memory in the stack is usually very fast compared to the overhead of heap management, even more when you have a garbage collector involved.

Neural Networks, Pre-Lenses, and Triple Tambara Modules

Neural networks are an example of composable systems, so it’s no surprise that they can be modeled in category theory, which is the ultimate science of composition. Moreover, the categorical ideas behind neural networks can be immediately implemented and tested in a programming language. In this post I will present the Haskell implementation of parametric lenses, generalize them to pre-lenses and introduce their profunctor representation. Using the profunctor representation I will build a working multi-layer perceptron.

Everything I know about SSDs

All HDDs, SSDs and flash drives have an internal controller. It's the way that the storage device can be, in the words of Microsoft, abstracted from the host. That abstraction is done by logical block addressing, where each cluster capable of being addressed on the storage device is known to the host by an ascending number (the LBA). The storage device controller maps that number to the sectors or pages on the device. To the host this mapping is constant - a cluster remains mapped to the same LBA until the host changes it. On an HDD this relationship is physical and fixed: in its simplist deconstruction an HDD controller just reads and writes whatever sectors the host asks it to. It doesn't have to think about what was there before, it just does what it's told and writes new data on top of the old. It does that because it can, there's nothing preventing a new cluster being written directly on top of the same sectors of an old one. On an SSD it's different.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Book of the Week

The Art of Invisibility: The World's Most Famous Hacker Teaches You How to Be Safe in the Age of Big Brother and Big Data

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

Hello again.

Comic of the week

Links of the week

The demise of coding is greatly exaggerated

Natural language is ambiguous and not suitable for programming. LLMs still need to generate code to get things done. If not inspected carefully, this incurs tech debt at monumental speed of the computers. The natural language prompts are not repeatable/deterministic, they are subject to breaking any time. This makes "natural language programming" unsuitable for even small sized projects, let alone medium to large projects.

Why does an extraneous build step make my Zig app 10x faster?

I had never thought too carefully about bash pipelines, but Andrew’s comment made me realize my mental model was wrong.

My mental model was that jobA would start and run to completion and then jobB would start with jobA’s output as its input.

It turns out that all commands in a bash pipeline start at the same time.

Sega Saturn Architecture

The Sega Saturn approach is what I consider a ‘semi-solved’ case. The VDP1 doesn’t implement any VSD function: You either feed the geometry in the correct order or you get a mess. However, Sega provided a graphics library called ‘SGL’ that implemented a solution called Z-sort or Painter’s algorithm [13] which performs polygon sorting by software.

Subscribe now

What Computers Cannot Do: The Consequences of Turing-Completeness

There is a difference between mathematical Turing-completeness and practical Turing-completeness, and for physical humans and machines, we should care about practical Turing-completeness.

Why? Because the only thing we should care about actual software is how good it is, and it is only as good as the lack of bugs and other defects.

Sure, our machines may not be mathematically Turing-complete, but if we still cannot prove properties about those programs, they may as well be Turing-complete.

The rule of thumb that I use: if you cannot prove even one substantial property of the program, it is Turing-complete.

The Real C++ Killers (Not You, Rust)

So the dear micro-optimizations we all love so much may both speed up our code by factor 3, and slow it down 90%. It all depends on the context. How wonderful it would be if a compiler could pick the best alternative for us so e. g. the index-sort would miraculously turn into swap-sort when we switch the build target. But a compiler can't possibly do that.

⏯ How To Handle Teammates Who Take All Responsibility But Do Substandard Work

This video is about how to handle teammates who take all responsibility but do subpar work. In the video, Jocko Willink suggests two ways to identify the teammate's intention: is it sincere helpfulness or a desire for power?

I have created a survey to get feedback from you. It only takes 2 minutes.

Go to Survey

Book of the Week

Grit: The Power of Passion and Perseverance

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

I have created a survey to get feedback from you. It takes only 2 minutes.

Go to Survey

Hello again.

Comic of the week

Links of the week

How Figma’s databases team lived to tell the scale

Given these tradeoffs, we began to explore building a horizontally sharded solution on top of our existing vertically partitioned RDS Postgres infrastructure. It didn’t make sense for our small team to re-implement a generic horizontally sharded relational database in-house; in doing so, we’d be competing with tools built by the likes of large open source communities or dedicated database vendors. However, because we were tailoring horizontal sharding to Figma’s specific architecture, we could get away with providing a much smaller feature set. For example, we chose not to support atomic cross-shard transactions because we could work around cross-shard transaction failures.

Unlocking Kafka's Potential: Tackling Tail Latency with eBPF

Slow produce requests without corresponding slow writes were always occurring around the time of some other slow write. We started wondering whether those requests were perhaps queuing and waiting for something to finish. By analyzing Kafka source code, we identified a couple of places that use synchronized blocks, including those guarding log file writes.

We set out to measure how much time Kafka’s threads, processing produce requests, spend on the aforementioned locks. Our goal was to correlate periods when they were waiting on locks with writes to the file system. We considered two approaches to do that.

The secret to luxury watch success: Tissot up. Tudor down. Why?

The Morgan Stanley report is in. It's not just numbers and statistics. For once this report gives a clear indication of what it takes to be a succesful luxury watch manufacturer.

Subscribe now

Building Columnar Compression for Large PostgreSQL Databases

By compressing your dataset, you can grow your PostgreSQL databases further. As we’ll see across this article, this highly effective compression design allows you to reduce the size of your large PostgreSQL tables up to 10-20x. You can store much more data on smaller disks (a.k.a. saving money) while improving query performance. Timescale compression is also fully mutable, making database management and operations easy: you can add, alter, and drop columns in compressed tables, and you can INSERT, UPDATE, and DELETE data directly.

Why Are (Most) Sofas So Bad?

The big change in shipping logistics came when furniture manufacturers developed new ways of deconstructing and flat-packing their product, allowing sofas to be shipped via regular carriers, both more cheaply and more quickly. This innovation opened up new avenues for penny-pinching. Favorable tax situations, cheap labor, and a surging wood and wood-type-product industry in Asia made it suddenly very inexpensive to construct furniture there, and those new shipping tactics made transportation feasible.

An Evidenced-Based Training Plan for Brazilian Jiu-Jitsu

Although inherently a self-defense system, structured BJJ competitions have increased in popularity and professionalism over the past 15 years (27), with elite international and numerous regional events available to the BJJ competitor. Additionally, many international grappling events offer lucrative prizes to higher-level competitors. The shift from self-defense to sport, whereby time limits and points govern victory, has increased the reliance on the physical capacities that were once only a minimal component of BJJ. As such, there is now an increased demand for evidenced-based strength and conditioning methods to provide competitors with a physical advantage.

Book of the Week

Extreme Ownership: How U.S. Navy SEALs Lead and Win

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.

I have created a survey to get feedback from you. It takes less than 2 minutes.

Go to Survey

Hello again.

Comic of the week

Links of the week

How I keep myself Alive using Golang

If something happens which causes my blood glucose to trend downwards, I wanted to be notified so that I can take an action before any more major issue appears. However, if things do take a turn for the worst or I’m unable to self-resolve, I wanted to feel good about the fact that someone else would get notified if I am unable to help myself and hopefully prevent it from escalating from a P3 to a P0 incident.

How video games use LUTs and how you can too

We’ll embark on a small journey, which will take us from simple things like turning grayscale footage into color, to creating limitless variations of blood-lusting zombies, with many interactive WebGL examples along the way, that you can try out with your own videos or webcam. Though this article uses WebGL, the techniques shown apply to any other graphics programming context, be it DirectX, OpenGL, Vulkan, game engines like Unity, or plain scientific data visualization.

The Race Car Faster Than F1

This is the Porsche that was faster than F1, and smashed a 35-year-old Nurburgring lap record… I’ve spoken about it before, but this time I got an interview with the Technical Project Lead - to understand exactly HOW Porsche did it.

And it’s crazy stuff, it’s almost the real-life version of my “F1 with no rules” video from a few weeks ago. They took their already, extremely fast LMP1 car, and made it over 12 seconds faster at Spa -- that’s 11%, a huge amount in motorsport.

Subscribe now

Nerdy internals of an Apple text editor

So to make typing snappier, I’ve built an algorithm that looks at the next character being typed as well as what characters are around it. The gist of the logic is that if you’re typing a special Markdown symbol, or the location of the edit is surrounded by one, then you should update the whole paragraph, otherwise you can simply rely on the typing attributes. It’s a simple algorithm that does marvels for the speed of the editor in the majority of typing situations.

The design and implementation of a lock-free ring-buffer with contiguous reservations

This is the story of how Andrea Lattuada (PhD student at ETH Zurich) and James Munns (from Ferrous Systems) designed and implemented (two versions!) of an high-perf lock-free ring-buffer for cross-thread communication. If any of those words look scary to you, don't fret, we'll explain everything from the basics.

Reverse Engineering Protobuf Definitions From Compiled Binaries

A few years ago I released protodump, a CLI for extracting full source protobuf definitions from compiled binaries (regardless of the target architecture). This can come in handy if you’re trying to reverse engineer an API used by a closed source binary, for instance. In this post I’ll explain how it works

Book of the Week

Becoming a Supple Leopard: The Ultimate Guide to Resolving Pain, Preventing Injury, and Optimizing Athletic Performance

Do you have any more links our community should read? Feel free to post them on the comments.

Have a nice week. 😉

Subscribe now

Have you read last week's post? Check the archive.