RE: Changing Subjects: IFML

["Joe Andrieu" <joe@andrieu.net>]

June 10, 2005 9:27 AM, Morbus Iff wrote:
===
Anyone familiar or have opinions on:

  Interactive Fiction Markup Language
  http://ifml.sourceforge.net/

It's an XML dialect for marking up interactive fiction. It seems like a 
dead project (last update 2002), the parser is in Java, not something I 
know anything about, and the XML is kinda "naive" (ie., telling the user 
to use [p] instead of <p> within a CDATA, etc.).

Have there been any other attempts at something like this? I found this 
along the way while investing QML, a Choose-your-own-Adventure dialect:

  http://www.questml.com/

I know the new Erasmatron's Deikto is going to be XML-based too.

Thoughts, comments, other examples?
====

Some sort of XML representation is definitely useful.  I'll have to look
into the references you mention to see what others are doing.

This is actually the area we (Jeff Rawlings and I) are currently focusing on
at Realtime Drama.  We've found that the XML schema is a good way to isolate
exactly what we are talking about when we discuss various aspects of our
system, without requiring the working code to demonstrate it.  

Unfortunately, when we talk to others--especially those in the game
industry--it can be very hard to communicate what we are doing. As one
friend put it, "It's like two twins with your own private language."  I'm
hoping that the XML will help distill some of our jargon into something that
can facilitate clearer conversations.

Interestingly, the questml project doesn't address the question of whether
or not their representation of the data-space is the right one. Fair enough,
that isn't its purpose. At the same time, I think that's one of the bigger
problems in interactive stories/interactive fiction: a profound
miscommunication and/or disagreement about the underlying *stuff* we are
working on.

I like IFML's use of scenes to define nodes of engagement, although the
"Dialog" functionality is fairly limited.  Two major things are missing for
me. First are temporal controls... so you can drive actions independent of
the stimulus/response loop of player input. Second is that the definition of
a scene appears to be more like a "room" as used on a MUD, rather than a
scene in a movie with a clear value transition to occur.

Nonetheless, both examples are concrete definitions of a type of interactive
story.

I'd very much like to see XML for stories where the story elements are
clearly abstracted from the presentation. That's what we are working on and
I must say it is slow going. Detangling the deep structure in a manner that
supports variability and interactivity is quite a challenge.


I'd love to hear about other examples.

-j

--
Joe Andrieu
joe@andrieu.net
+1(805) 898-9389