Whither Seltani

I am reminded by Carolyn’s comments about multiplayer Twine in another thread…

So last year I built an engine for multiplayer choice-based text IF. It’s optimized for people to build small environments and text game experiences, and share them in an online space (with MUD-style chat features).

What do I do with this thing?


I originally built it, or rather set up the one demo server, for the Myst Online fan community. Most of the discussion has been on a Myst fan forum (guildofwriters.org) – not because I thought this forum was inappropriate for a hypertext MUD, but because that crowd got excited about it. There was about three months of activity from that direction – not all of Myst fandom, but a handful of interested people – and then it tailed off.

Seltani is not Twine. The multiplayer aspect changes the assumptions, which affects the requirements, which affects the technology, and the result can’t be simply described as “Twine with multiplayer”. On the other hand, whenever somebody asks about multiplayer Twine, I want to talk about Seltani as a possible technology.

I feel like there must be some audience for this thing. I don’t know what it is. I’ve demoed it at various small game events in Boston, but nobody’s come rushing up to me saying “This is exactly what I’ve been looking for!” And yet, MUDs exist and Twine exists and fan writing communities exist.

(Formally “Tworld” is the open-source engine and “Seltani” is the Myst-themed server I’m running. I don’t think labels are the issue here, though.)

So what?

The one thing that kept Seltani from being “exactly what I’ve been looking for!” is that it was specifically about Myst fandom, rather than a community-defined world. :slight_smile:

Okay, I have heard that from some people. I always regarded the Myst theme as a proof-of-concept and a gift to the Myst community to whatever extent they wanted to make use of it.

On the other hand, I don’t have a better hook off the top of my head.

To be clear: the engine is theme-neutral. Out of the (github) box, the UI talks about “portals” rather than “linking books” and you can tweak that as you want.

When I picture an ideal multiplayer Twine experience, my core focus would be on making Twine-style games that two or more people play together - sort of like the Portal 2 co-op experience of interactive fiction. I picture a pick up and play experience that goes as smoothly as starting a regular Twine game.

You load a game-specific URL, which asks: “Are you starting a new game of [game title] or joining an existing one?”

If you say you’re starting a new one, then you get a join code to hand out to other players. If you say you’re joining an existing game, then you get a prompt for your join code.

You play through either synchronously (where the interface pauses until it gets both/all responses, and then loads the next screen) or asynchronously (where earlier players affect later players, but not vice versa) or some mix between the two.

Seltani didn’t seem like a good match for this to me, both because of the Myst overlay and because of the required registration. That isn’t to say Seltani’s not awesome (I think it was an amazing accomplishment!) but it doesn’t feel right for my needs.

I was really excited when I first saw Seltani, and I still think it’s an awesome project. “This is exactly what I’ve been looking for!” is actually what I thought at first sight. And I wanted to test it myself so I tried to build Tworld it but it was quite complicated at that time (the package available for my distribution were too old). Now I think I’ll have a look again…

About Seltani, I don’t have any knowledge in Myst, but it looks interesting. For enjoying it better, maybe I’d need to play some Myst chapters…

Thanks, that’s a helpful viewpoint.

When I started designing Seltani, I began with the desire for a consistent and persistent world. (A set of instanced, but consistent, worlds with persistent state.) So logging in was a requirement. Later I added the ideas of a guest account (one-click temporary login) and a URL entry path (so you don’t have to come in through the site’s front page). But I haven’t thought about works that are supposed to stand completely on their own.

That concept doesn’t fit well with the Myst Online theme (which has a clear intro-room and a clear hub area). But a different setup could have a mythology which is less closely bound. It would not be hard to extend the engine to support that. (Although there are several missing pieces to build.)

That leaves the problem of building a work. The current setup is biased towards environments with state. You are talking about constructing in synchronous or asynchronous pages. You can do that (it’s a scripting language, you can do whatever) but you might find that it’s more work than you hoped.

On a separate topic: social features. Right now the front page lets you see if other people are logged in and what worlds they’re in. And you can chat with other people in the same room (and instance). That’s it.

I always figured the Right Path would include personal buddy lists, and also a guild/team/group structure. So you could direct messages to specific friends or to a team-chat channel. Obviously I never got around to building this.

(Teams could also jointly own instances. That would let them set up shared bulletin boards and private spaces.)

Do you think this would be a big boost to your use of the system? It’s running counter to the “platform for independent work” viewpoint, but I expect there’s room for both.

Not for me.

If I wanted to make something more like a MUD - a broad, arching experience - then likely yes. But I’m looking at tight, short experiences - say, half an hour or so - and I don’t really want the players to be distracted by things outside each other and the game they’re playing. Breaks mimesis, such as it is.

cvaneseltine, i was talking about something v similar yesterday, i agree completely. MUDs overshoot my idea of complexity–short and minimal is better. i want simple online Twine ahhhh

Would it be possible for you to make…I guess like a “side door” for Seltani that was not Myst-themed? More of a generic multiverse lobby so people might not be put off by the D’ni trappings. You could still link from there to the Myst neighborhood and have a list of Myst worlds, but something that might not jar so much if it led to a futuristic cyberpunk instance with laser guns.

Perhaps something like a movie theater multiplex common area, or perhaps a sleek lobby with an elevator to different worlds?

I like that you can have direct url access right to a particular world, that is nice. Also that someone can log in as a guest without being registered.

How customizable is a Seltani world? Can you do any HTML or CSS trickery? (was a looooong time since I looked in.) Could a Seltani world link a player to a separate Twine and back again? In that case, you could have, you know, say “Porpentine’s Universe” where she has a neat little MUD that leads to all her games.

i dont MUD…i SLIME…


I would rather run two separate servers with different configurations.

Like I said, there are pieces missing. (Guest accounts are not properly wiped between uses, which means that you can’t really use them for state-based works. But that’s fixable, if I ever have the wherewithal to do more Tworld development.)

Currently no. The world is displaying itself inside the Tworld interface (web) app, and that puts tight constraints on what trickery is available. (Open Javascript access => malicious world code hacking your account => death of kittens.)

This is one of the big ways in which Tworld Is Not Twine.

But the current server is even more strict than that. I consciously ruled out all author style control beyond bold/italic/Myst-font, because I wanted that consistent mythology – and a consistent UI experience to go with it. This is something that could reasonably be server-configurable… except that you still run into Javascript security limits pretty darn quickly.

A platform for truly independent, Twine-level-configurable multiplayer worlds is a new engineering problem. Mind you, I’d love to work on that if there were a paycheck in it. :slight_smile:

The thing is that I’m kind of deeply interested in the community aspects of this stuff! It’s the MUD experience that attracted me to this project in the first place. Having people hanging out online and jamming new story ideas, and then being able to walk into the commons room and open a portal and say “Here’s my new thing!” And then you have groups branching off and spinning up their new concepts and collecting their favorites and building art galleries and so on.

(This thread might as well go in the “Choice-based IF” section, now that we’ve got it.)

It seems like the idea here is a heavily non-persistent platform, more like FPS deathmatch than MMORPG. So many security concerns would be relieved by the fact that there’s no real need for people to have accounts.

That’s a good point. I hadn’t thought about that.

However, what people are talking about does require some persistent state – short-term rather than long-term. So from the server’s point of view, it’s not entirely different. But it might still be possible to eliminate accounts.

Of course, some authors will start with short-term persistence and then wonder why long-term persistence is impossible! It seems like an obvious small step. I’d really want a range of options, and making that scale up nicely could be tricky.


I think (hope?) the confusion could be largely avoided by choice of terminology, since there are lots of types of games that have only short-term persistence. On the other hand, completely doing away with accounts might make it impossible to have slower-paced games that take place over days or weeks.

…maybe something with frames?