Hi Patrons!
Thank you all so much for your continued support of my game development! I thought I’d write up some notes for you about the development process of They Will Not Return. This post contains spoilers for the game, so stop reading now if you haven’t played it yet and not being spoiled is important for you.
Origin
The game was inspired by the science-fictional idea of robots living on after humans go extinct, as seen in works such as the Russian animated short There Will Come Soft Rains (based on a Ray Bradbury story) and this series of covers by Mel Hunter for F&SF magazine. In the first part of the game you’re basically playing as the house robot from There Will Come Soft Rains.
Unlike Rage Quest: Disciple of Peace last year, which I was already half way through making when I decided to enter it, I wrote They Will Not Return as an IFComp entry from the start. This didn’t affect the content of the game much–IFComp games are very varied, so there isn’t a particular style you need to aim for to do well–although I did write it with the assumption that most players would only play through once, so I didn’t put in the same level of multiple endings that Rage Quest has. There are a few scenes you only get if you make certain decisions, but most of the differences are details rather than entirely different paths.
Changes during development
I was originally going to have the game track stats, like the Rage and Discipline meters in Rage Quest, and I started to implement it that way, but when I saw it working I decided that they weren’t appropriate for this game. It’s about deciding what to do in different situations, not building a character in a RPG-type way. So I cut out the stats and just tracked individual decisions.
Another idea I thought about was having the player pick a type of robot to play at the start (house robot, military robot, maybe medical robot, etc.) and have a different first chapter for each of them, before the stories converged. I think this would have been a good idea if the game was a lot longer (novel-length rather than short story-length), but with the scale of game I was going for I realised it would take up too great a chunk of my development time, and it would be better to spend the time making one more fleshed-out opening chapter.
Finally, the focus of the story shifted as I wrote it from being very heavily about the relationship between the player character and QT, to being more generally about what you do in the post-apocalyptic setting, with sending QT away and trying to have nothing to do with them being a valid option. For that reason I decided to cut the subtitle from the original title (They Will Not Return: A Post-Apocalyptic Robot Love Story) and just call the game They Will Not Return.
Visual style and presentation
Because it’s a game about robots, I decided to re-use the visual style I had used for Industrial Accident (white text on black, with grey icons on the left, and links highlighted by putting them in boxes). In fact, if you like linking things together, it’s possible to read They Will Not Return as a loose sequel to Industrial Accident, although I didn’t want to announce it as such for fear that people would think they needed to play one game to understand the other.
Outside of the game, the cover art, title, and blurb are areas where I learned from last year’s IFComp entry. Several of the reviews of Rage Quest criticised the cover art and title, and to be honest I’m kind of embarrassed about the cover art now. With TWNR, I put more effort into making cover art that would fit with the other covers on the IFComp ballot (again using the same technique as I had in Industrial Accident: taking a stock image from Pixabay, tweaking the colours, and putting the title over it). I also went for a snappy one-sentence blurb of the kind that a lot of other games use for good effect, rather than the verbose book cover-type blurb I used last year.
Switching to Tweego
This game marks a switch in my development process from using the Twine IDE to using Tweego, a command-line compiler for Twine stories. I’m already using this system for Beyond the Chiron Gate, but TWNR is the first game I’ve completed using it. With Tweego, you write the code for a game in any text processor (in my case Notepad++), rather than being locked in to the Twine IDE. You lose Twine’s visual map of how the passages connect, but you gain a lot of flexibility in how you work on your code.
I found that switching to Tweego changed how I approached writing the story, but I don’t know whether that had an effect on the end result. With Tweego, for example, you can put multiple Twine passages in the same text file, so it’s easier to write lots of small passages and link them together in complicated ways that would lead to an unwieldy tangle in Twine’s story map. On the other hand, it makes keeping track of the story structure the author’s responsibility, and if you’re not careful it’s easier for dead ends to slip in.
Release and entry
I found I was pushing right up against the deadline for competition entry, and when I finished the game I felt like it was good enough to enter but perhaps not as polished as I’d like. The ending, in particular, might feel a little rushed, and I didn’t have time to have many people playtest it so there are probably weak areas or bugs that a more thorough playtest could have caught. With that in mind, I’ve been surprised at how positive the first few reviews have been, so I’m glad I finished it and entered it.
Apart from fixing bugs, I’m not planning to make changes to TWNR, as I’d rather get back to work on Beyond the Chiron Gate. TWNR will remain playable on the IFComp site until 15 November, at which point I’ll move it to my own site.
John