I need a project. What’s a narrative tool project that nobody’s done that you would use? Some missing Godot plugin? Dialogue authoring spreadsheet? Interactive comic framework?
Bonus points if you’d pay me money for it, obviously, but that’s not a requirement.
(I asked this last year and wound up with the Visible Zorker project, which I’m still pretty proud of. So all wild suggestions are welcome.)
I wrote a response on my own mastodon profile, but with federation working the way it does, it felt prudent to bring my response to the forum, as well:
The thing that I would selfishly want to see realized is a visualization tool for Inform 7’s Scenes. This doesn’t lend itself to visual representation as cleanly as Rooms do, and many games would have no benefit for such a tool, since either they don’t use scenes, or the scenes themselves do not interact and are wholly independent.
However, I have a baseless suspicion that scenes are underutilized by the community and represent an extremely powerful tool which we could be using a lot more inventively than we have been.
If I could see relationships between my scenes the way the index lets me see relationships between my rooms, this could make it a lot easier to troubleshoot my games. I think it would also lead me to attempt to create more reactive environments and game-states.
I imagine the wider community can come up with better suggestions than this, though, so what’s your answer? What’s a narrative tool that you’d like to see created?
A wish I’ve seen more than once from Twine authors, and which in a different context would help for any dialogue system (including my own hand-rolled for Inform) is some better way to visualise branches.
I’ve seen Twine authors say the visualisation of the branches (in that 2d physical space of the screen) simply gets big and hard to track, and there is no particular or better way to deal with it.
My analogue in Inform is I’m writing lots of dialogue scenes, and there is no broadly satisfying way to manage their branches except by manual scrutiny and thought. I write them in a vertical document with numbers, and bits saying ‘go to 2’ etc. It’s the same issue, except I’m probably dealing with fewer nodes than Twine folk, but mechanically it’s still awkward.
I can’t recall ever seeing Choice Of authors saying this kind of thing, but perhaps they say it on their own forum, not here.
Talking aloud, maybe I should see if I can bring Inform’s Skein to bear on my own problem. But the problem there is, for that to work, the content already has to be implemented in Inform. And I design/write most of this first outside Inform.
If I think of some kind of non-existent tool, it would be visual, like Twine, but more manipulable. Maybe you could tag paths and light them up, or jump-zoom to clusters or divergences or convergences, or highlight these different types of things. And you could click a node to just copy its text so you could paste it over into your program of choice. Or maybe export all the node text if you prefer. To my knowledge this is not a thing that exists?
Anyway, I suppose if You also have had thoughts along these lines, speak up. It can’t just be for me. I’ll survive
Yes, I think visualization in Inform would make many things easier to manage. It would also be nice to make things that are more visual. And perhaps the answer to that is just “Use Vorple.” I want to push back against that, though. Are visual experiences exotic in 2025? Nowadays, if I ask myself “what do I like about Inform,” I don’t really think about the command prompt. I think about the world model, rules-based programming with natural language, and so forth.
In the proposal for Dialog Sections (Inform 11), there is a mention of a theoretical (and external to core Inform) p&c interface running on top of it.
But I agree with Wade, I try to imagine (and it’s just imagining, I haven’t seen anything more than the proposal) trying to do something like Portrait with Wolf with Dialog Sections, and the long, vertical document does seem hard to manage.
My solution in Portrait with Wolf was to break everything into small tables (there are 150 tables in PWW!), then break those into 13 extensions so that I could arrange them all on my screen. This practice led me to temporarily abandon the Inform IDE for VS Code, even though its Inform extension isn’t 100% compatible with Inform 10.1.2. I think the game would have been observably worse if I didn’t own a wide-screen monitor.
I agree that scenes could be better understood. I’ve written about them for that reason. I think that the skill floor for starting with the documentation is higher than maybe it has to be. We’ve seen beginning authors say so! Perhaps there’s a tutorial/documentation opportunity for scenes and other features (relations come to mind). Better visualiztion of the plot would be welcome, too, of course.
EDIT & who knows, maybe interfacing with or learning from other systems is a way to accomplish some objectives (perhaps some sort of twine interface)
Interactive comics is kinda how I envisioned Poink-and-Clink being used, although I’ve sorta stalled on developing it, in part because it turns out I don’t really have ideas for little interactive comics? But it might be worth checking out if you’re interested in that sort of thing.
I’m interested in finding suggestions for accessibility-related projects, either for myself or my students. Any big gaps that need to be filled, whether it’s in tooling or playing?
I’ve been thinking about accessibility a lot lately! I specifically considered the IFTF’s Accessibility recommendations while making both Portrait with Wolf and Marbles, D, and the Sinister Spotlight. I think some important suggestions there involve providing access to information, attempting to make potential actions visible (status bar in MDSS or the Motif listing in PWW), providing geographical information.
I’ve always said–though maybe it hasn’t come across–that a parser game “story” mode is an accessibility feature.
I’m not saying that I’m doing these things well. An important question would be if there are “best practices” for a system or even extensions/models for approaching the committee’s suggestions.
Note that a lot of this isn’t really parser game specific, which I think is a strength of the report.
Yeah, that’s what I’m looking for. A project that can be of general use across games, interpreters, tools, whatever you like. Think of something like extending a glk implementation to better support accessibility. I was thinking of adding TTS and STT directly to a glk, but I get the impression that OS features are really doing a good enough job of this; correct me if I’m wrong
(This approach is really something that’s not possible with Inform itself, since ideas submitted to the evolutionary process are really property of Graham Nelson thereafter AFAICT.)
I’m very skeptical of “visualization” as a technique for understanding branching narratives (or code).
Visual aids can only really represent a few dozen things before they become as complicated as the thing you were trying to understand in the first place.
And when analyzing messy node diagrams, it’s not just the nodes we’re trying to visualize, but the lines connecting the nodes (the “edges”). We can only visualize a few dozen of those, and that typically means we can visualize only a handful of nodes at a time.
Visualization only works in trivial examples where you don’t need it; it fails in complex environments where you need it the most.
Instead, I claim that you want automated tests.
Are there no more sections marked “TODO” here?
Ensure that there’s a path from X to Y
Ensure that there’s no path from X to Y
Ensure that all the paths from X to Y meet certain criteria:
All paths from X to Y pass through Z
You can’t get from X to Y unless certain variables are high enough
It’s not that I want to just look at 36 nodes. I can already do that in Twine. I want to see and follow things about them, and probably crucially, overlapping things about them. That’s where I need a better tool.
I thought more about my idea last night and I was thinking of a tag system. You could apply tags to nodes. Now you’ve got a list of tags. Click one and all with that tag light up. Maybe there’s now a scale tool. It’ll zoom to the first cluster. If the path is linear enough you could just tap something to move along it, at your preferred scale. If it’s not you could tap auto-numbered branches to follow the cluster that way.
If there’s a keystone moment with its own tag, you could now select two tags, and see if the path from the second tag hits that moment. At the most specific level, you could tag all of one path to pull it from the list if you needed, though obviously if you did this for everything you’d just replicate the choice list and there’d be too much info.
I think something like this, plus good export options (e.g. being able to click a node to grab all its text - Twine is already not good for this kind of thing if you’ve tried it as an aid to another sytem) would be a boon.
-Wade
PS I was just thinking, if Roblox wasn’t one of the worst systems in the world for handling text, maybe I could even program this myself in Roblox It could do all the other stuff.
I readied this interesting draft (along others; thanks for the link !) and I think is a dialogue system consistent with I10’s natural language (and by first reading looks like that implementing it shouldn’t introduce new incompatibilities…)
It could do all the other stuff.