A 7-minute techno live set created entirely in GLSL shaders that fits in just 64KB. Yes, 64kb. This WebGL intro by 0b5vr was submitted to the Revision 2023 demoscene competition. Procedural visuals meets algorave meets extreme compression. My mind is blown.

I just released dgtools, a command line utility to work with the Discogs data dumps. This little endeavor was supposed to be a quick side quest, but it transformed into a rabbit hole.

Discogs is the go-to service for record collectors. They might have one of the biggest databases for physical music releases. On a monthly basis, they release a compressed XML of a subset of their database under a CC0 license. Tools already exist to import them into a PostgreSQL database, but I wanted the flexibility of a custom-built solution. I started building something in a Ruby on Rails app but quickly diverged to Go as I didn't want to pay the ActiveRecord performance cost.

I have quite a bit of experience with Go. I started learning it in 2012, and some code of mine still runs on thousands of production servers worldwide. I haven't really touched the language for a few years though. So it was interesting to write a small project and learn a few things.

Go's XML parser is great to work with. With a mix of streaming and parsing into tagged structs, it is very convenient and performant. I wasn't expecting it would be SO easy to reliably parse XML.

Go's ecosystem is rich. I thought pq was the go-to choice when working with PostgreSQL and Go, but the development has stalled. I hit a bug where pq is unable to bulk load data into jsonb columns. pgx is now the right choice for that kind of workload. Similarly, there are tons of good libraries when it comes to structuring a CLI application. And don't get me started on the cool TUI libs!

I forgot how amazing Go's embed feature is. Embedding migration files into a binary is so easy and such a useful feature. The ability to distribute a single binary is my favorite feature of Go.

While Go correctly produces and consumes multi-line CSV, PostgreSQL cannot import them. Unless I missed something, having new lines in a CSV means that PostgreSQL cannot process them correctly with a COPY FROM 'filename' query.

Go is a framework as much as it is a language. The documentation, standard library, and bundled tools make it an amazing built-in developer experience for small-scale projects like the one I wrote.

Late to the party (as I often can be), I recently discovered Rosa Menkman’s work while at NXT Museum for the “Still Processing” exhibition.

Turns out, Rosa Menkman has quite the background in Glitch Art, having worked on theorizing it and having produced artworks bought in the Stedelijk Museum collection. Usually not a big fan of video essays, I ended up being very interested in two of her productions. The first one about racial and sexist biases in analog and digital image processing. The second one about the changing nature of rainbows due to atmospheric conditions (pollution) or changes in our analog wetware (eyes).

Glitch Art isn’t something I have ever been passionate about. Rosa Menkman’s work managed to ring a bell with me though.

Her work on DE/CALIBRATION ARMY (2017) exposes how test cards were used in photography or cinematography to provide standardized colors. She also reveals how the same standards were applied in digital image processing,  exposing the elephant in the room: sexist and racial bias in image processing. I found it very interesting. 

Commoditized as it is, we tend to forget about how much we live in a world of computational imagery. We are aware of the algorithm in our social networks, but do we really often think about the algorithms in our image capture and processing stack? Not so sure.

Capturing an image with a smartphone involves a deeply integrated technological stack of hardware components (lens, sensors, chips) and software (processing algorithms, analysis, correction), each maintained by teams with their own requirements and constraints (time, engineering, financial, etc.). 

Anyone who has ever worked in engineering or product teams knows this fundamental truth: requirements necessitate choices.

To choose is to renounce. Biases love choices because they thrive on renouncement. Choices compound through the whole technological stack. Choices are then amplified by the default settings which are mostly defined by anticipated market expectations. Biases are now ready for their reinforcement loops until they get disrupted by other forces. A new externality has been created and needs to be addressed. 

Rosa Menkman demonstrated this through her glitch art. She created an app you can use to apply biases to your portrait. The result’s fun and glitchy-looking sure, but in no way different than beauty filters used by social media platforms such as TikTok (search for Bold Glamour Body Dysmorphia on your favorite search engine). These beauty filters are created to sustain and improve engagement within these apps. In that case, our relationship to biases is weaponized for maximizing profits. 

Choices... Choices...  

Through the small lens with which I see her art, Menkman exposes the challenges coming with the embedded power structures of our everyday technologies. 

Are you old enough that you spent your time on IRC back in the days? Do you like your chat infrastructure free of billionaire influence? Do you have free time? Are you experiencing nostalgia? If you answered yes to one or more of the questions above, here’s what it takes to connect to the IRC network in 2025.

The IRCv3 protocol update aims to fix the (let’s be honest) awful user experience of IRC. They even have a draft for server-side chat history! Meanwhile, the Open-Source community congregates on libera.chat, and if you're looking for channels on different networks, Netsplit.de is a great resource. 

Bouncers act as a middleman between your client and the IRC server. Installed on a server, they help with keeping chat history and hiding your home IP address.

ZNC works well and is acceptably complicated to set up. The good folks at LinuxServer.io even provide a docker image. There's also Soju, an alternative written in Go but I felt lazy and didn’t try it. 

Did you know that mIRC is regularly updated (and can you imagine that a IRC software has a 4 letter .com domain)? Their licensing model hasn’t changed. It makes me happy that not all software has turned to a subscription model. If you're looking for a multi-platform client, Halloy is written in Rust. It’s a nice piece of software. I like the UI, and the icon is gorgeous.

I don't know why anyone would be on IRC in 2025, to be honest. It’s clear that IRC has a worse user experience than proprietary solutions such as Discord, Slack, or Telegram. That said, IRC is open and has a following strong enough that 30 year-old software keeps being updated and developers are building new open-source clients in memory-safe languages.

Using IRC isn’t easy, and might never be. And it’s a quiet place. But it’s free of corporate interests and sometimes I think that’s the best feature of all. And quiet is nice. 

P.S: I’m marc_in_space on libera.chat.