Tech: Blog

Machine Learning requires vast amounts of resources, and distributing these resources across multiple devices and regions helps with cost, speed, and data sovereignty. Bacalhau is an open-source distributed orchestration framework designed to bring compute resources to the data where and when you want, drastically reducing latency and resource overhead. Instead of moving large datasets around networks, Bacalhau makes it easy to execute jobs close to the data’s location, reducing latency and resource overhead.

The move to the cloud promised to save users money and give them insights into their usage and costs. However, the opposite happened. A 2025 report from AAG stated that around 82% of respondents found cloud spending challenging. A cloudzero report from 2024 states that more than 20% of respondents had no clear idea of their cloud costs, with reports for large users sometimes consisting of thousands of rows of hard-to-read usage data.

Welcome to a new newsletter/post/radar/term-yet-to-be-defined from me that I have been planning for ages. I intend it to be something of a round up of tools and services i’ve been trying recently, plus also industry analysis and trends. To begin with it’s just the tools and service round up as the analysis part requires more thought that I haven’t had time for yet.

Autocomplete has existed on macOS in some form for years, and it's always been inconsistent. I tried Cotypist, an AI-powered autocomplete tool that works across almost every app on the Mac, to see if it does a better job.

With many applications that rely on data warehouses, you need to keep data sources in different locations. This could be due to privacy or regulatory reasons or because you want to keep processing close to the source. However, there are still times when you want to perform analysis on and across these data sources from one location but not move the data. This post uses Bacalhau to orchestrate the distributed processing and DuckDB to provide the SQL storage and querying capacity for some mock sales data based in the EU and the US.

When the modern-day internet began emerging in the early 2000s, finding hosting services and resources to run the new wave of dynamic web applications was hard. You needed a database to store application data. These were slow, expensive, and unreliable, regularly bringing applications to a grinding halt when a single instance failed. You needed a server to run interpreted languages like PHP, Python, or Ruby. These were equally expensive, often needed configuration, had security issues, and frequently ran out of memory or CPU resources, again bringing applications to a grinding halt.

A while ago I made a video on Mermaid, the markdown-style syntax for creating diagrams and charts. It's still one of my more popular videos, and it needs an update. In the meantime, the people at Mermaid Chart, a new commercial company formed by some of Mermaid's original creators, reached out to show me their platform. So in this post, I take a look at what it adds on top of the default Mermaid syntax.

For those of you in the Bazel ecosystem, when you hear "Dagger", you probably think of the dependency injection framework for Java, Kotlin, and Android. And for good reason. Dependencies are frequently part of a build process. But there's another Dagger in the wild, a tool and platform for ephemerally building and testing multi-language projects. Sound familiar? In this post, I look at what (the new) Dagger is, how it works, and how it compares to Bazel.

Changes in recent versions of macOS have made running macOS and Linux virtual machines far easier and more performant. Primarily used for testing, a plethora of new applications that take advantage of these features emerged, so when Mac Stadium released a free desktop version of their long-running Okra range of virtualisation tools, one feature in particular stood out to me as something different and potentially useful.

I've been a mostly happy VSCode user for a few years now. So much so that I even maintain a couple of extensions for it. Before VSCode, I loved Atom. And before that… I can't remember anymore. But every now and then, it's good to look into alternatives and see how they compare, what options you haven't considered they offer, and if they bring anything new to your productivity.

Ruby’s creator, Yukihiro Matsumoto (Matz), released the first public version of the programming language in December 1995, making Ruby just shy of its 30th birthday. It spread across Japanese-language Usenet newsgroups, a popular way of exchanging conversation and media before the World Wide Web, and then reached broader communities throughout the late 1990s.

Developers talk a lot about “immutability.” Outside the technical world, it usually means something negative: unmoving, inflexible, and entrenched. However, in the technical field, these features become an advantage and mean that a system is reliable and reproducible.

I've been happily using [Parallels Desktop][1] for my Windows and Linux virtual machine needs for about the past three years. It's performant, supports a reasonable variety of guest operating systems, and, with some tweaks, can also run headless developer machines and Kubernetes clusters.

Installing and updating applications and other dependencies on a computer really should be a solved problem by now. Yet almost every major desktop operating system provides multiple options, with no real clear answer to “which is best.”

JetBrains is known for its variety of developer-focused IDEs, all based around its IntelliJ core. In recent years, they developed their own writer-focussed plugin, Grazie, for handling grammar and style guidance, Writerside for a wide variety of tech writing tasks, and more recently, an AI assistant for writers. The company has also helped with community-maintained plugins such as those for markdown and asciidoc. But then, a year or so ago, JetBrains announced that Writerside would become a fully-fledged tech writing IDE. I've been using it on and off for some time, and in this post, I take a look at some of its features, recent updates, and related plugins.

A few years ago I detailed the hardware and software I use for producing my podcasts and videos. Since then, I have changed a lot, so it seemed a good time to revisit.

In June, Kubernetes celebrates its tenth birthday. The system is now so widely used by hundreds of thousands of companies worldwide to scale their applications to meet demand it’s hard even to remember a time before it existed. But there was a time when other options were available, and I even remember using some of them.

Who do you think reads most of your technical writing? Customers? Potential customers?

Open source is at a crossroads. For the past few years, venture capital has directly or indirectly paid for many of the contributors and much of the infrastructure it needed to keep going.

How many words do you write every day? Think about it. Every time you send an email, post in Slack, or comment on an Instagram post. Oh, you thought I would mention those wondrous words you pour into your latest novel, blog post, marketing copy, or script?

I create a lot of content with audio. Interviews that become final podcasts, final podcasts, videos, and more. In almost all cases, a transcript of that audio is useful. Often, I use it to create a blog post based on that audio, and I am also in the process of creating interactive transcripts for my backlog of podcasts. And that aside, accompanying transcriptions of shows are good for SEO and accessibility.

We had to put our 17-year-old cat to sleep last week. First, why does English have such terrible words for this action? “Putting down”, “putting to sleep”. They are both terrible ways to describe such an action, but I’m not sure I have any better ideas…

Markdown may be popular for its simplicity and "plainness", but it's often full of information. Whether for internal or external consumption, that information often contains code snippets for running certain commands or code examples. Wouldn't it be great if you could run that code directly as you read from within the markdown?

I've mentioned before that I have been using Obsidian more and more, and from a knowledge management perspective, it's fantastic, with a plethora of plugins that extend its functionality to connect to many external services. But as a cross-platform Electron application, integration with the host operating system is one of the places it fails. As the underlying data store is a folder of files, some integration is possible there, but Obsidian has minimal awareness of changes that happen.

I have a past life in music. In the late 1990s, I produced one of the first online fanzines and wrote for several street press magazines in London and Melbourne. In the middle of that, I also played in bands in the UK and Australia, playing guitar and, latterly, drums. And then I stopped. My musical creative spark had run out. Fast forward to about 2021, and it returned, primarily due to relearning the instruments I had always played by “feeling” in the past. Actively learning instruments and production techniques familiar to me properly opened my musical mind again, and I am slowly producing music I always wanted to make in the past before joining rock bands.

I am extremely lucky to have a display problem. I have a small home office and a studio setup. At home, I sometimes have my laptop under my screen and sometimes to the left. In the studio, I have two external screens, one of which I switch to 1080p when recording videos. So that’s four potential screen configurations. I have always found screen configuration on macOS relatively smart and reliable. It tends to remember individual screens and automatically switch to the configuration you last used with that screen. But if you want to manually change layouts or resolutions, what do you do? You can keep jumping into system settings to make changes, but that gets tedious quickly. I wondered if there was a better way.

What do American Online, the Apple Newton, DVDs, Tamagochis, MP3, PDF, and Sony Discmans all share in common? They were all products born in (and some dying in) the 1990s. An era I remember vaguely well as I spent most of my mid to late teenage years in it. And now I see the fashions and band T-shirts from that era back on the streets and…

It’s been a few years since I attended Berlin’s Internationale Funkausstellung, better known as IFA. And after several slow years for conferences and trade fairs, capped off with a slowdown in tech and tech-adjacent spending, I was interested to know what the show floor would feel like.

About 18 months ago, I covered my attempts to replace Evernote. I was surprised by how popular that post was and how much feedback it received. Unsurprisingly, I have adapted and changed my setup since then and in this post, I cover what I now use and how I use it. Hopefully, with the same level of readers and feedback 😁.

In this post, I cover how I use Aeon Timeline, an application for building interactive narrative timelines. If you want to read more, then take a look at the post I wrote on the Setapp applications I use regularly.