This is a brief post sketching out a synthetic style of cryptography. In this style, one doesn’t appeal to any kind of complexity theory, probability, or even adversary.

In which I ramble about large language models.

Recently, I’ve been working on a threshold ECDSA protocol—and its implementation—called Cait-Sith. I thought I might write a little blog post to talk about it.

The typical presentation of theoretical cryptography has one central goal: defining what it means for cryptographic objects to be “secure”. I think this goal is misguided.

This is a short post on distributed key generation (DKG) in the context of cryptographic groups, such as elliptic curves.

I’ve been playing around with STARKs lately, meandering my way towards a toy implementation.

Let’s talk about one of my favorite cryptography papers.

In this post, I’d like to provide a technical introduction to key encapsulation mechanisms (KEMs), with a focus on proving the security of various constructions.

Many topics in cryptography on this blog so far, but not many basic topics. This post is a crack at providing such an approach. With luck, it should bring utility to unfamiliar folk, but also grins for folk familiar with this art.

Many cryptographic protocols attempt to satisfy a notion of “identifiable abort”, where if a malicious party causes the protocol to prematurely halt, then they can be detected. In practice, I think that this notion isn’t all that useful.

This blog post is an introduction to state-separable proofs, a technique for proving the security of Cryptographic schemes.

This is just a brief post going over a few books on Cryptography I’ve read, and would potentially recommend to people interested in the topic.

The protocol so fun you have to implement it! Like I did recently.

When proving security, Cryptographers often model hash functions as random oracles, which act like random functions. In practice, hash functions are different from random oracles. The question is: does this difference impact security?

Secret sharing seed phrases, studying MPC, and Yao’s Garbled Circuits.

It turns out that Go v1.18’s generics functionality is enough to encode traits, which has many applications. In particular, we could write a generic library for Elliptic Curves, among other things.

A toy blockchain in Haskell, simulation-based security, and tinkering on Cryptographic foundations.

Monero is a cryptocurrency which claims to be “private, and decentralized”. One of Monero’s main tools towards this privacy is the ring signature. Ring signatures allow you to sign on behalf of a group, without revealing which member of the group you are. They can be constructed as an elegant extension of Schnorr signatures, and aren’t all that hard to understand either.

The ECDSA signature scheme is quite ubiquitous, used everywhere from TLS to various cryptocurrencies like Bitcoin. Funnily enough, it turns out that it suffers from a few malleability issues, although I doubt these pose a serious issue in practice.

Recently, I ported my NES emulator, Ludus to run in the browser, using WASM. You can play around with it here. This post is a brief overview of the interesting aspects in creating this port. Ludus Ludus was an NES emulator that I wrote 3 years ago, back in 2018. I was starting my BSc at EPFL then, so it’s a bit fitting to revisit the project now that I’m starting my MSc.

Twitter’s Bluesky initiative created a little challenge where the goal was to verifiably link different digital identities together. This is my attempt at this.

How do Bitcoin’s new Taproot signatures interact with the good old key derivation methods from BIP-32? It turns out that the answer isn’t all that straightforward.

Designing a hash function where the order of inputs doesn’t matter is surprisingly easy.

I’ve developed a few intuitions about algorithms on quantum computers. This post is an attempt to share them.

It turns out that all you need to make a signature scheme is a way to prove your identity.

Recently, I made a toy E2E encrypted messanger, called Nuntius. I had fun tinkering on it, and thought that some of the cryptography involved would be fun to explain.

End-to-end encryption is a very appealing guarantee of privacy, and more applications want to provide this guarantee. Web applications are popular, and they want to implement this functionality in the browser. What kind of guarantees does a user still have with a web app, served to them dynamically?

Recently, I’ve been working on a little encryption tool called Nimotsu. My goal with this project was to implement all of the cryptographic primitives involved. I had a lot of fun doing so, and thought it would make for an interesting blog post.

Over the past couple months, I’ve been working on a library for constant-time Big Numbers in Go. I think it’s about time that I presented a bit of this work.

This is just a quick post, crystallizing some of the ideas I’ve had recently about organizing numeric classes in Haskell.

Last week, I made a little web application for visualizing some fractals, and I thought I’d write up a few thoughts about how it works.

Recently, I’ve been experimenting with using spaced repetition for self-studying “advanced” mathematics. This post goes through my motivations for adopting this system, as well as a few techniques I’ve used in adapting it to mathematics.

This is just a quick post about a programming problem that’s been circulating around on twitter recently:

Tychonoff’s theorem proves that the product (even infinite) of compact spaces is also compact. The proof makes judicious use of Zorn’s lemma. In fact, it uses it so well, that I gained an appreciation for how fun the lemma can be.

I’ve been thinking recently about handling strings and names in compilers in a more principled way, and I think I’ve come up with a nice way of doing this.

I was reading a cryptography textbook, and I came across an algorithm that reminded me why I often dislike descriptions of algorithms in textbooks. This is a short post going over this example, and thinking about how to improve these descriptions.

Recently I’ve been wrapping up my work on a compiler for a subset of Haskell. While my subset doesn’t (yet) have any support for IO, I’ve been thinking about how to implement it.

This is a post going over the basic concept of defining objects through Universal Properties, in Category Theory, with explanations and examples in Haskell.

This is the third post in the Haskell in Haskell series.

In this post, we’ll go over creating a parser for our subset of Haskell. This stage of the compiler is responsible for taking the tokens that our lexer produced in the previous part.

This is a quick post about the Chinese Remainder Theorem. Specifically, how to use it to solve a system of system of simple modular equations.

This is the second post in the Haskell in Haskell series.

In this post, we’ll go over creating a lexer for our subset of Haskell. This stage of the compiler goes from the raw characters in our a source file, to a more structured stream of tokens.

We’ll finally be digging into some serious code this time, so get your editor ready if you’re following along!

One cool thing about lists is that they have canonical ways of consuming and producing them: folds, and unfolds. It turns out that these are canonical, in that folding and unfolding functions are themselves isomorphic to lists. In this post, we’ll explore why this is true.

This morning I decided to implement my own version of UNIX’s classic tree command in Rust. I thought the algorithm was both non-trivial, and interesting enough to warrant a quick blog post.

This is the first “real” post in the Haskell in Haskell series.

In this post we’ll go over setting up a basic project in Haskell. We’ll be building our compiler on top of this foundation in the next posts.

This post is about the different ways objects can be immutable and mutable in programming languages, why deep immutability is what we really want, and a path towards implementing that in higher level languages.

This is an introduction to a series I’m calling Haskell in Haskell. The goal of this series of posts is to go through the implementation of a subset of the Haskell language, from parsing, to typechecking, to code generation, using Haskell itself.

If you want to find out how many bytes a certain portion of a struct uses, you might be tempted to do some arithmetic with sizeof, but this will yield unpredictable results!

When you write a parser in Haskell, you want to use parser combinators. When you write a lexer, you should use lexer combinators! We’ll see what these are, and how to use them to write a simple lexer.

This is about some shower thoughts I had recently about the infamous Monty Hall problem. Namely, how to make sense of the counter-intuitive results involved. We’ll see how reasoning counterfactually can make the best strategy seem a lot clearer.

Having recently inaugurated the 4th version of this blog, I want to go over the technologies that I used to make this, as well as some of the tricky things I had to do to make things work.

This post is a basic introduction to the idea of Categorical Graphs, or just the standard theory of Graphs, developed through the lens of Category Theory, and generalized in the obvious ways that follow through that lens. This is just an introduction mainly because this theory doesn’t seem to have been developed very far yet, and I haven’t been able to develop it that much independently so far. I think I have a good grasp on some very basic ideas here, and wanted to present them while they’re still fresh in my head.

So this is going to be a bit of a Rant on some things I’ve been thinking about recently. To set the background for this post, I’ve been working on a note-taking app with a focus on individual blocks referencing eachother. Kind of like Roam. The idea is that this kind of system can be used for discrete nodes of knowledge, which we can call blocks, for simplicity. Blocks don’t contain much on their own, besides some text, and references to other blocks.

So, recently I’ve started working on a text editor, using Gtk, via gtkmm in C++. I’ve decided to try my hand at writing the text area widget from scratch, just using the text layout facilities provided by Gtk’s Pango. My rationale for this is twofold: It seems that Gtk’s TextView widget is missing a few nice things, like source lines, and using the more developed SourceView is not doing a lot of the work myself.

Recently (well, more like a month ago), I came across this interesting observation: In my head, I immediately jumped to the notion of Algebras in Category Theory. I had recently studied that notion, found it quite interesting, and was very happy to see this observation, because it was actually quite obvious to me thanks to what I’d recently learned. The goal of this post is to unpack that tweet, and then explain why that observation is true.

In this post, we’ll cover 3 ways I know of encoding the natural numbers $\mathbb{N}$ in your standard functional language with recursive types and polymorphism. At least, these are the 3 most generalized ways of doing it. As we’ll see, some common encodings are just specific cases of a more general encoding. Prerequisites Some familiarity with defining data types in a functional-esque language might be helpful, but shouldn’t be strictly necessary.

The usual definition of a Group excludes empty groups by definition. There are alternate definitions of a Group that allow us to include the empty set, and which are quivalent to the normal definition in all other cases. This post explores this alternate definition, and the resulting differences with the normal concept of a Group. What is a Group? A Group is one of the most important structures in abstract algebra.

The concepts of monomorphism and epimorphism are very important in Category Theory. I always had a hard time remembering which one was which until I thought about a good mnemonic.

Recently I made an app for making group video calls. The difference between this and something like zoom is that there’s no central server responsible for routing video calls between the participants. Instead, I used WebRTC in order to set up peer-to-peer (P2P) calls between all of the members of a group.

This post is about how to parse programming languages which define blocks using indentation, instead of braces and semicolons. We’ll end up learning how to infer these blocks based on the indentation, using Typescript to parse a toy functional language.

I read a lot of stuff, mainly on the internet. One problem I’ve always had is keeping track of the important bits of the things that I read. The most important information sticks with you, but it’s not easy to remember everything. There’s always a nagging fear that there’s some importing bit you might be missing, or that you might forget something useful.

This is a post about how I work with data in fullstack development. Specifically, I share what I think are good patterns for sharing data and logic between the frontend and the backend of an application.

This is a review of The New York Trilogy written by Paul Auster in 1987.

As the title says, this book is actually a trilogy of three shorter novels: City of Glass, Ghosts, and The Locked Room.

So I’ve been tinkering around with better systems for taking notes recently. I’m not a big fan of keeping notes, around because I don’t find myself actually revisiting them, and when I do, I don’t get much value out of it. On the other hand, my experience with language learning has definitely showed me how effective spaced repetition is when it comes to keeping things in your head.

This is a quick post about a “gotcha” I encountered recently in a React application. This involved the use of React’s useState hook, which had a subtle difference between how I thought the hook worked, and how it actually worked.

When it comes to language learning, one approach that I really subscribe to is immersion. This means trying to absorb as much of the language as possible, as often as possible. It’s often said that “speaking with locals” is a good way to accelerate the learning process, and there’s a lot of truth with that statement.

This is a post about one of my favorite aspects of networking: how different protocols and concepts are layered together. I go over the concept of Layerability and how it applies to networking and programming more generally.

This is a post about Poline, a tiny programming language I wrote recently. The main “gimmick” of Poline is a feature called Green Threads. In fact, Poline doesn’t have many other features besides them.

This is a post about how different languages handle the concept of interfaces. We’ll go over the classical OO way of handling them, with Java, to the more recent approaches of languages like Rust, as well as others in between.

This is a post about ginkou, a tool I made recently. This tool uses Rust, SQLite, as well as mecab to archive sentences, and then to retrieve them based on the words they contain.

This is a quick post about the difference between Data Races and Race Conditions, and how data structures or patterns providing freedom from data races can fail to provide race condition freedom.

This is a short post about a crate I recently published: https://crates.io/crates/ludus. This crate provides the core logic of an NES emulator, and can be used to build independent GUI applications.

This post details a useful pattern for organizing stateful components in functional code. This post assumes knowledge of Haskell, up to Monad-Transformers.

Having worked on a bittorrent client in Haskell recently, I wanted to write about a few tricky aspects of the implementation that I encountered: managing the concurrent piece downloads, and saving them to files.

When programming in a Functional style, Mutability is often avoided on principle. But what’s more important is having functions be immutable from the outside. In this post we’ll see how to use interior mutation while providing an immutable interface.