David Stark / Zarkonnen

Before Airships: Conquer the Skies, I spent several years working on another project, Space Exploration: Serpens Sector. This was a game about exploration and crew management heavily inspired by the first Strange Adventures in Infinite Space. I put a lot of work into it, but the result wasn't really satisfying because I had neglected to create a good core game design. So I built up a complex UI and complex encounters too early, and each change became a lot of work. Eventually, I rebooted the development and started on a new, simpler prototype that concentrated on the core mechanics.

Then Airships came along, a "side project" that pretty much instantly got more traction and interest than SE:SS. Within a few months, it became my main project. I was tired of going around and around with SE:SS and wanted something else. SE:SS was never exactly cancelled, just put way on the back-burner. This was four years ago.

The other day I re-downloaded the source of the first SE:SS version and got it up and running, and I found myself pleasantly surprised. Yes, the core mechanics are probably ultimately not satisfying. Yes, it's overcomplicated in places. But it runs, it's perfectly pleasant to use, and there's quite a bit of nice content. The pre-rendered graphics look kind of dated, but in a way that can plausibly pass as charming rather than hideous.

So why not release it as is?

After a protracted beta, Airships version 9.2.4 is out, replacing the old multiplayer system.

I recently came across this Mitchell and Webb classic: "Cheesoid", the story of one man attempting to make a robot that can smell things.

Apart from being pretty funny, I think it's a great rapid-fire illustration of a lot of common mistakes in machine learning. That's what Cheesoid is, after all - it's a classifier system for smells.

This post was originally published in 2013, on the previous incarnation of this website.

I just read a pretty neat article that argues for more interesting forms of magic in computer games. One idea I really like (because I'm horrible) is having to pay a permanent price for each spell cast. (The spells, in exchange, being very powerful and useful.)

At around 1 AM, with a solid two hours to go before the jam deadline, I uploaded and published Dryad, my 3rd Ludum Dare entry.

It's a platformer about a forest creature wishing to become human, and, uh, killing a whole bunch of humans in the process. As you become more human, your magical powers wane, and you become slower, weaker, less perceptive, making each successive level harder. You can choose the order in which you lose your powers, and some orders are easier than others.

Alongside small flying units like planes and air hussars, I originally planned to add small ground troops capable of directly attacking things. It would be the obvious final recombination: there are flying units that directly shoot targets (planes), flying units that board targets (air dragoons), and ground units that board targets (marines). But there are no land troops that shoot targets.

During development of dev 9.2, I realised that there were major obstacles to making ground troops work, which is why I dropped them. Here is why:

After a record number of beta iterations, Airships v9.2 is out! It introduces small flying units and troops:

The giant bee is a prime example of the influence that Suspendium can have on nature. Most bees are small, fairly harmless animals, but this particular species has learned how to incorporate Suspendium in its shell. As a result, these creatures can grow to massive size.

A bestiary entry about giant bees.

Their floating hives appear seemingly overnight. A floating bronze capsule with a giant mainspring in its centre. Man-sized mechanical wasps issue forth from it, harassing people and livestock, dismantling machinery where they find it. Each wasp is its own intricate clockwork held aloft by suspendium, powered by a spring that takes up most of its central body. Even so, the wasps have to frequently return to their hive to be re-wound by the mainspring, or they risk becoming sluggish and eventually inert.

A bestiary entry about clockwork wasps.

Apart from improving boarders, Airships 9.2 also adds flying troops. This dev blog entry is about about the details of their code.

In Annulus, you guide a stone-age tribe living on a ring-shaped space station. Apart from hunting and gathering, there are strange technological artefacts to study, and on occasion, there's an opportunity to shape the culture of your tribe.

The next version of Airships will focus on improvements to troops. In this post, I'm going to write about the performance and pathfinding problems that large numbers of boarding troops face, and how they have been resolved for the next version.

I live in Zurich with my partner and two cats. My main current project is Airships: Conquer the Skies, a Steampunk shipbuilding RTS. I also do smaller prototypes and games on occasion, as well as maintaining Selenium Builder, participating in Open Data work, and doing freelance software development.

If you're looking for game development advice or feedback, I'm happy to help.

My free time is largely spent playing board and computer games and cuddling the aforementioned partner and cats. I also organize game development meetups, do talks, fiddle with various art styles, and write about stuff. I have a morbid fascination with all the bits of history that get left out in school. I'd like to believe that rationality and kindness are not opposed.

Airships now includes instructions on how to set up your own multiplayer server. Here's why this matters:

I was recently thinking about just how satisfying fancy-looking progress bars in games are. When you've got a nice, swirly XP bar, and it fills up, and you gain a level, that feels viscerally great. Meanwhile, progress bars on the web tend to be minimalist. So I decided to make some decidedly fancy progress bars using CSS.

When you're creating a new ship design, it's nice to be able to work quickly. Having to scroll through the list of modules is pretty tedious, which is why Airships has long had search for modules. In version 9.1, I've improved the search system, allowing you to get the exact module you want with very few keystrokes.

A quick pixel piece from this photo reference.

I've been thinking about the different kinds of grids of tiles and blocks used in games, which come in a variety of scales and shapes.

The HMS Sojourner was an experimental airship from some years back. A huge Suspendium crystal had been found in a deep mine, and the kingdom's best scientists had examined it and found it to be flawless. A crystal of this size, powered up, could lift a ship to unprecedented heights.

The Sojourner was built around the crystal, carefully cradling it in a giant frame of shock absorbers. Huge boilers and generators provided the power to energise the crystal. A team of experienced engineers tended to the engines. The captain was a steadfast veteran.

I realised that I was sick of futzing around with Google Slides for simple presentations, so I made a plain text to HTML slides compiler. There are many like it, but this is mine. I called it Talkify.

I like writing about bugs in Airships. I don't want to present myself as some infallible rock-star Indie developer, because I'm anything but. Airships is a game for builders and tinkerers, and I have seen again and again that you like reading about its creation, warts and all.

So today we delve into The Mystery of HMS Longcat.

A lot of the work I'm putting in for the next version of Airships relates to its interface. Be that improved module search in the editor, or the ability to mod in new map sizes and difficulty levels. One particular improvement that's long overdue is the ability to scale the user interface elements depending on the screen size.

I just bought and played Voyageur, a science fiction story game for mobile. It's one of the games supported through Failbetter's Fundbetter programme, and since I'm mildly obsessed with Sunless Sea, I thought I'd give it a try. I was largely disappointed, though.

I won't bore you with explaining how my own unreleased SF story game would be totally superior. You've read it here before, and comparing a real and finished game with a set of ideas is hardly fair. Instead, let's dig into what makes Voyageur unsatisfying.

I've been ill over the last few days, probably as an inevitable consequence of doing so many events in January. So I've been playing a fair amount of Sunless Sea, a great game if you're ill, because it's quite soothing in a "dark sea filled with monsters" kind of way.

And as always when I play Sunless Sea, my thoughts turn back to Space Exploration: Serpens Sector, my old game project that got backburnered hard when Airships: Conquer the Skies took off. The two games are in the same "go from port to port in your ship and do stuff" genre, after all.

You are in a spaceport, and your Mars rocket is ready to board.

This game is dedicated to Elon Musk, who very much wants to go to Mars, but won't stand up for people just trying to get home.