top expert

a version history of RTE.

(open spoilers follow)

I worked on the Spring Thing version of Repeat the Ending for a total of eighteen months. That doesn’t include the time I spent on its earliest incarnation as a novel. It had a rocky test history. I blame myself and my failure to accurately describe the game to testers. I got feedback like “too many nouns” and “too many verbs.” It was a text game with too many parts of speech! A couple of testers heavily implied that I was wasting their time.

However confident I may have seemed, I was an insecure beginner after the fashion of insecure beginners, and I was very discouraged by the experience. I was lucky to have found a couple of very kind people who believed in the project. Without them, there would be no Repeat the Ending. It is all because of them, and I mean this sincerely. This is something for a tester to remember: we never know what the author feels or thinks about their work, nor can we know the differences we might make with our comments.

It was fairly common for testers to… disappear. Based on conversations at the Interactive Fiction Community Forum, I don’t think this is common. I may have been hard to work with. Perhaps the game was too dark. I don’t blame anyone. Sometimes things just aren’t a match, you know?

I’m an iterative writer who is used to beginning at the center before radiating outward. This is my process for writing a poetry collection, and I am a poetry MFA. I took a similar approach to Repeat the Ending. The first, proof-of-concept version involved the opening segment as well as an incomplete presentation of the central problem of the trailer (get out).

The second version introduced the first version of the scoring system, which dictated the ending that the player got. It also introduced themes central to the game. It was still early days: there was no Battle Princess Chiyo. There was no way to check the player’s current goal, either, which must have left some testers confused.

In the third iteration, the trailer park was created, which reinforced some other key themes. I also introduced the footnotes, which were originally intended as a tutorial only. I had so much fun writing them, and a couple of testers liked reading them, too, so the idea of a scholarly text was born. That’s a key part of my process: letting the idea grow, being willing to follow along. This may surprise you, by my writing is a lot like free climbing. I’m improvising and looking for holds. I’m trying to find my way. I write to discover the text, to surprise myself.

Then, I was a long time away. I wrote the rest of the game, in terms of plot and puzzles. Updates included solidifying the four critical voices commenting on the work, all the while building up the idea of 1996 and 2003 versions of the text. I wrote the emotional ending that included, of all things, my appearance as a character. In his honor, the current release of the game is credited to Drew “Drew Cook” Cook. More testing, and, with it, a realization that the critical edition concept must be expanded to include essays and other documentation. I investigated options before settling upon Wade Clarke’s menu and help menu extensions over Emily Short’s. Why? Wade had a screen reader mode.

The next version was more polished and included a complete paratext of reviews, snippets, and so forth. More testing. It really felt like things were shaping up. A push to a final, Spring Thing version followed. Lovely artwork by Callie Smith was added to the game. I wrote the “infamous 2003 transcript” and coupled it with Callie’s demonic “fashion guide” for the orange-eyed woman of the prologue.

With only days before the Spring Thing deadline, I began the process of creating a web page out of my Inform game–a necessary convenience for many players. This proved to be quite a nightmare. While Inform 7’s documentation makes no mention of this, it is apparently known that its web publishing function suppresses display of all images. A shadowy journey into python scripting and relative pathing and who knows what else ended with a website that–speaking frankly–I found rather unattractive. A desperate battle with cascading style sheets, a technology that I had never thought about or looked at, ensued. Nothing worked as I expected. I was very frustrated, honestly, that an eighteen-month project led to this. In the end, I was able to get colors and font looking the way I wanted, though the layout–complete with frames–had a vintage feel.

I wanted to switch from the QUIXE to Parchment interpreter. Multiple testers disliked the way Lectrote and QUIXE handled the ESCAPE key while reading the paratext. That didn’t have anything to do with my code, but I wanted to publish the web page in Parchment in response to their feedback. Parchment had different CSS, I learned. After some time fixing that, I learned that Parchment’s autosave feature bricked RTE play sessions. The developer fixed the problem quickly (for which I was grateful), but I was under the gun. QUIXE it was.

With Spring Thing came some constructive feedback. The hints regarding the score were considered bad, and people were right to say so. They would need to be improved. Despite very thorough proofreading, typos and miscellaneous glitches remained. There was an accessibility improvement. I had all of these suggestions implemented for a post-festival release.

The many reviews and ratings that Spring Thing received blew me away. Nothing had prepared me for its positive reception. I had put so much into it, in terms of both time and effort, and it felt amazing to have it praised and recognized.

Besides making the switch to parchment, Version 2 included an “after you’ve seen it all” section in the hints, which corresponded to the old “for your amusement” section in Infocom’s Invisiclues booklets. The update was short-lived: perhaps two hours after it was submitted to the IF Archive, I received a bizarrely hostile message about the opening passage, which had remained the same through Spring Thing and actual months of testing. At the time, I thought: I don’t want this person to be mad at my game. I changed it and put R3 out within an hour.

I regret that, now, changing something for them. I’m not sure that they even played the game.

Version 3 was around for a while. It was only different in that single respect. A version 3.6 followed. Its primary goal was consolidation. Instead of maintaining two separate repositories for the game, I redirected all IFDB traffic to itch.io, and began to use a single compile for both web and local versions. That’s been great. I can update anytime I want and let my IF Archive submissions trail behind, just in case bugs are reported.

Itch has some site usage statistics, too. I know some people find those figures discouraging, but getting just a handful of weekly visitors feels like a win for my strange game.

enough history (r4).

Version 4 of Repeat the Ending, just released today, is hopefully the final release. It is primarily a feature update. As I already mentioned, the guide to earning all of the game’s 33 points was simply inadequate. In fact, it was just a list of actions to perform. This section was a bizarre outlier in a work so obviously happy to be featuring hints.

Just look at this shabby embarrassment!

Because many of these points are only attainable at very specific points in the walkthrough, they were hard to clue. Hopefully, I’ve found a way. I also loosened the requirements for what I call the “true” ending, the one in which the protagonist meets his creator. My hope is that players can arrive at this ending organically as a matter of course, provided they have been blessed with curiosity.

This time, I used the new online single-file webpage generator for Parchment. After the struggles I’ve had with site creation, this feature is a very welcome change! It also honors light/dark OS preferences, which is a better option than configuring a single one manually. I’m very happy to cease my own clumsy attempts at web publication.

I do have to admit to peeking in to change the font. I really dislike serif fonts! However, this was a very simple change, one that even I felt equipped to make. It’s my hope that this will be the final release of Repeat the Ending, though we’ll have to see how my new Story Mode extension holds up under a bit of scrutiny. “Story Mode?” you ask, “What’s that?” Story mode is my first and only attempt (so far) at writing an extension for Inform 7. If you’re like me and don’t know a lot about programming stuff: an extension is a bit of portable code that can be incorporated into a separate program. My goal was to create a story mode not only for me, but for anyone who might want to use it.

I really enjoyed the experience, and I’m eager to tell you about it. In fact, I want to say so much about it that it deserves a post of its own. Once I get the code uploaded to GitHub, You’ll be the first to know. See you then!

next.

Whenever I get the story mode extension uploaded (I’m waiting on a dependency), I’ll explore that in detail. If it takes a while, I may start a dev diary for my next little project. It’s a children’s story for the Spring Thing back garden, built from the ground up with story mode in mind.