First feedback on my WIP, very positive, which I’m sharing not to boast, but to express, if need be, that a methodology that is not at all organic, but rigorously planned, is possible with Inform and its rules (vs. objects) orientation. I don’t know much about programming, but I do know a fair bit about creating worlds and plots, so perhaps this half-naive explanation will be of interest to someone.
I’ve mentioned here and there an approach, not very original in game design in general, consisting of encoding a set of world-model rule(s), closely linked to the fictional universe, then the world setting (the spatial conformation of the world, its lore), and finally the story, the tip of the iceberg in contact with the player’s imagination. In this order.
Why go through all this? Because in my case, I’m embarking with a great deal of pretension on a series of successive complete stories, with a metaplot and progressively more complex lore. This means being able to reuse code not strictly linked to the story, maintaining it separately (versioning, improving) and limiting the use of third-party extensions to keep the code under control during major changes (including the new Inform release, which I’m told is scheduled for January 2025 - French humour, sorry).
During the transition to the third phase of the cycle (the story itself), I experienced 2 weeks of this: it was a bit like the moment of truth.
Now, not everything is perfect of course, there are still things that need to be improved to be (automatically) told in a more fluid, more natural way (although sometimes I find that the slightly mechanical side of the parser is part of its charm). Sometimes (often!) I’ll have to make do with something that works quite well for the time being, rather than something that could be better but doesn’t yet work very well. And then I’ll have to spend some time customising the default responses and hardening the code against circumvention attempts (following the advice given in this thead, for the moment in French, but with @stormi 's permission, I’ll derive a list of checks in English during January). The aim is to start playtesting in the beginning of February (I’m lucky: I know several English-speaking volunteers personally). But the main thing I want to say about my experience is this: it’s working. It’s really working. For several days now, each code session has been 90% story, description and plot, and 10% technical. And that 10% of technical stuff reinforces the idea I had of my codebase for the world model: closed to modification but open to evolution. And that’s really cool! Because my work now is to concentrate on the most enjoyable part of the project, and I know I’ll be ready for the Spring Thing! Of course, I don’t claim to equal the best authors, but it’s cool to offer something that comes from our imagination to an audience that’s curious about other people’s imaginations…