I'm just back from the North American tour which ended at Phrontisterion, a microconference in Chris Crawford's forest lair in Oregon. Chris compiles reports from the participants - I'll link to that when it's up - but I wanted to make a couple of observations.

Chris is the Don Quixote of computer-mediated narrative. I'm not being denigratory- it's a role he chose himself. It's very hard to talk about the man and his work without sounding either credulous or aggressively sceptical. He was a major game designer in the eighties but has spent twenty years in self-imposed exile from the industry, trying to build his own vision of interactive storytelling. He has an extremely specific idea of what that should be - a strongly procedural environment populated by semi-autonomous characters who act on human-controlled characters, on each other, and on objects in the environment. Something like the Sims with all the resource poured into narrative sophistication rather than spatial simulation and good visuals. (In fact Chris is - to my mind rightly - very scornful of a spatial focus in a narrative context.)

If you think that sounds cool, I agree. If you think it sounds like a tall order for one guy in the wilderness, I agree. Chris has recently talked about how he considers the latest version of the technology a failure, and what he plans to do with it next. It's hard for me to talk about that without resorting to platitudes - I Wish There Were More People Like Chris. He Serves As An Inspiration. Interesting Failures Advance The Art. Here's To The Crazy Ones, etc. etc. I hope he produces something the world will find useful; I hope he looks more closely at other existing approaches to the same end; I don't get the sense he'll ever be satisfied with what he makes. I am profoundly glad he's still trying to make it, and a lot of what he says is still fiercely relevant.

What it brought home to me - and this is something that Dan Fabulich was very firm on - is that more than most art forms[1], interactive narrative is a series of compromises between the possible and the practical. An author has a limited amount of time to write a story before they have to feed the cat or go to work or run out of money or die, and there are only so many words you can get on paper (and remove from paper, polish, rework) in that time. You make choices about what to leave out as much as what to put in. When you make a work interactive, it's like going from a two-dimensional to a three-dimensional version of the same problem. Every possibility with a choice attached implies extra complication and sophistication. It absolutely requires ruthlessness or focused artistic intention about which outcomes you want to model and which consequences you want to reflect - not just the results of choices, but the emergent combinations of those results.

So there very quickly comes a point at which frugality is not only a virtue but a requirement, and you have to answer players' requests for more content with 'sorry, the game doesn't go there' [2]. But there is also a big win for using an existing platform whose grain or texture or toolset draw your consequence-modelling in one direction or another - in other words, impose some of your authorial trimming for you. The IF community and the visual novel community both benefit from having a body of expertise and tradition around approaches to their particular tools. We want to achieve something very similar with StoryNexus, but I would hate to see a monoculture - IF has benefited from an ongoing dialogue and competitive development between different tools. I am very energised by the recent rush of tools around text-heavy choice-based stories - StoryNexus, Varytale, ChoiceScript, Inklewriter and so on - and the differences in texture between those approaches will allow the specific benefits of each to shine, in a way they wouldn't if they were each the only option.

And the same goes for the larger context of more procedural versus more scripted tools. My creative approach - hence Failbetter's - sits firmly on the 'heavily scripted, gently procedural' portion of the continuum. Right now we model stories and story elements, not characters and objects. But it's essential for us to keep peering over the fences into neighbouring territories, both to find approaches and techniques we can borrow, and to better understand the limitations and advantages of our own design choices.

Which is a good impetus to talk about some of the reasoning behind design choices in StoryNexus; and I'll be doing that next.

[1] My take on 'are games art' in seven words: 'Yes, but generally not very good art.'

[2] or 'sure, but we'll charge 5 Fate for it' or indeed 'good idea, here's our Kickstarter'