Populating the Reach: EXOGENESIS

By Failbetter, May 23, 2017 · Alpha Sunless Skies

Our EXOGENESIS sprint is our first specific pass at populating the Reach, which means working on its denizens from officers to the agents you’ll encounter in the skies. Our art team, which has recently doubled, continued their work on ports, brought some agents to life, and created our first pass art pass over the UI. Content also continued work on various ports, while also creating a magnificent spreadsheet that will help keep track of Sunless Skies’ officers. Meanwhile, tech busied themselves with expanding on the pathfinding discussed last sprint, along with implementing tooltips (hugely important for our wordy games), balancing survival mechanics, and implementing agents.



While Paul is busy tackling Titania, one of our more complex ports so far, we’ve wasted no time putting Toby to work. His first job was Hybras, a rather fungal settlement with a dark secret.


Toby also created one of our agents, a Tackety Scout, while Paul tackled the Scrive Spinster. During the Sprint Review we all oohed and awed over both, though the Tackety Scout stole our hearts. In James’ words, “It looks like it might explode at any minute….it’s perfect.


Meanwhile the art team has been giving Liam’s initial UI wireframes an Art Nouveau-influenced facelift. Below you can see some initial designs on how they might look, though we’re sure they will go through a few iterations.



The writers tackled two more ports this sprint. James fleshed out the London stronghold, Port Prosper, where London Loyalists (or opportunists) can accept bounties for hunting pirates and Tacketies. Meanwhile Cash designed the Leadbeater & Stainrod Nature Reserve, where a research outpost conducts investigations into the secrets of the Reach.


Another beautiful spreadsheet has also been created, this time it’s all about officers. While the writers are currently pitching their content ideas for officers, this spreadsheet details which slot they’d go in (First Officer, Signaller, Quartermaster, Chief Engineer, and

- of course – Mascot), where players would recruit them, which bonuses they’d provide, and how bonuses would change when upgraded. This lets us ensure that officers provide an even distribution of bonuses, and ensure that there are satisfying choices to be made for each slot.

Non-Player Locomotives

Meanwhile, we’ve started turning our Kickstarter backer survey responses into locomotive names! We’d like to note that no fewer than three people recommended “HML Ubiquitous” and we feel strongly that all three should be entered in the game.

In Sunless Skies, when certain non-player locomotives spawn in the game, they’re given unique names. So instead of seeing a generic Tackety Scout, you might see the ‘UCE Vagabond’. These names are drawn from a list of six hundred entries inspired by our wonderful backers. Our task has been to take information provided by them and turn it into names that fit the period and tone of Sunless Skies. Six hundred times. Six actual hundred.

If you haven’t filled out your Kickstarter Survey, please do so at your earliest convenience! And if you know someone who hasn’t, poke them!


Continuing the pathfinding work for combat, Barry has focused on long-distance pathfinding this sprint.


 Some background on this: pathfinding in Sunless Skies will be based on nodes, points on a mesh. When an AI agent needs to get somewhere, it looks for the best (usually, the shortest) path from its current location to the destination node.

The world of Sunless Skies is very large, and pathfinding between nodes that are very far apart is computationally expensive. We could solve this problem by making it so that AI agents only move around when the player is nearby. But in Sunless Skies we want agents to sometimes remain active and pursue their own agendas at distant locations. So we need a more efficient method for them to find their way around.

The solution we’re using is to group all the nodes in a given area together, and treat them all as a single point on a larger meta-mesh that covers the entire map. Even though it covers the whole map, the number of nodes on it is manageable. So the AI can use it efficiently for long-distance pathfinding, and then switch to local pathfinding to deal with obstacles in its immediate area.


Our next FBG Podcast will be this Friday 26 May, at 16:00 BST! This week will be focused on art, to answer any questions you may have about the art process, especially now that we have multiple artists!

No Comments

Leave a reply