You're asking about the parts of the system. The noteworthy are:
* a STRIPS planner for the NPCs in the game that has to deal with stochastic actions
* an NLG subsystem
* all the usual game OOP stuff that would work with the above (immutable objects with semantics)
* a layered storytelling framework
But if you're asking about the 7 years, it's mostly a lot of pivots. The layered storytelling framework was an idea that I had for a very different game last year, for example.
If I knew exactly what I'm building, it would take about 6 months of work, even with my limited free time. But that's not accounting for all the things I had to learn-by-doing along the way.
No no no, you go big or you go home. You can't be a little bit The Eye of Argon.
I meant that it has a lot of parts that don't work well togther: the presentation doesn't match the writing, the writing doesn't match the setting, the slots doesn't evoke a feeling of a contested combat. It feels like a lot of ideas thrown together without an overall vision tying them together.
I keep thinging of Inkle's Sorcery! and how well it's put together, or Simogo's Device 6's novel use of text.
> a STRIPS planner
Well now! I'd love to hear more about this; I didn't notice it in action while running backwards and forwards from the war forges to the corpse room & back while waiting to see if the explicitly-mentioned time-pressure was really a thing or a unconvincing plot momentum device.
What are the parts of this system, and what were the blockers? The game is turgid, and the buttons are too big (the unchosen options should disappear on each decision point).