Kent... I'm appalled. Ball sucking jokes are FUNNY? I'm sorry, I just don't get the humor. Every day, people suck balls and are caught by their mothers and fathers and it scars them for LIFE. For you to toss around ball sucking jokes like you think the prospect of this is funny makes me so horrified that I want to lay in the fetal position in my shower for hours until the pain is washed washed washed away.

I suggest you think of the ramifications of your "ball sucking" jokes before you go and type out (what you think is) funny quips about it. If "balls" meant "typing posts that wound" you would be the ball suckingest ball sucker this side of Ballsuck, Indiana.

Oh, and the new manual is still on the list--it's just that it's a big list.

Good question about what a "world model" actually is. I've had this discussion with a few people since you asked, and a good working idea of what a world model entails probably revolves around exactly what a programmer has to do in order to make certain things happen in a game: i.e., how the player is implemented, how actions are interpreted, what the default behavior of the game is (meaning, probably, what the library does) for a given situation and what opportunity there is to customize that behavior, etc.

In that sense, Hugo is more similar to Inform in terms of world model. There is a "physical" representation of containment (using the object tree, although Hugo doesn't have to depend on it) and behavior is mostly governed globally by pre-existing library routines, with opportunities for the player to intercept default behavior . (In Hugo, the means of intercepting behavior are somewhat more flexible. At least that's my probably biased take.) This differs somewhat from TADS, which has a more object-oriented approach to default behavior using methods on the individual classes and objects. (If you're familiar with C and C++, TADS is closer in many ways to C++ than Inform or Hugo are. But then again they all support objects, classes, multiple inheritance, and the ability to affect default processing is tied to objects/classes, and so on.)

(Written in a rush, so hopefully I didn't confuse anyone, including myself.)

Back in late 1999, early 2000 I wanted to make a little text game with some pictures -- I wasn't good enough to get anything along those lines working without having access to existing code that did what I wanted to do, though. Kent had released the source to Guilty Bastards by that time, and I was able to pick out the graphics code from his source and adapt it so I'd be able to do the same.

I remember thinking when I first played Guilty Bastards, "Hey, this seems like a Magnetic Scrolls game!"

(Plus, the syntax for Hugo is similar to Inform, so I was able to kind of hit the ground running in that respect.)

And Kent, how's the new version of the Hugo Manual coming along, and is it still supposed to be published eventually by the IFLibrary?

I think the manual is still being in the process of being edited...? I will wear an ear to ear grin when that is ready. Presuming it's the same cost neighborhood as the Inform one I want to have one for home, one for the office and one to take to bed with me, like a teddy bear, only with much greater advice and input for common programming questions than the average bear.

I suppose my question was nebulous because I really have no idea what people are talking about when they refer to a 'world model'. Is it essentially how the library handles the various types of objects?

Just out of curiosity, ICJ, why did you switch from Inform to Hugo?

And Kent, how's the new version of the Hugo Manual coming along, and is it still supposed to be published eventually by the IFLibrary?

Yeah, it really does depend what you're referring to. On a certain level, their world models are are very similiar: in all the major languages, you create programmatic objects which are "physical" objects in the game world; rooms are just one type of object, characters are another, boxes and keys and vehicles are others, etc. It's then the function of the library to manage all these objects in a simulationist way, and that's where the difference lies, and all comes down to different methodology, levels of implementation, and complexity.

What kind of things are you looking for, specifically? Personally, when I switched from Inform to Hugo, everything more or less "worked" in the same manner, with the big exception that trying "examine wall" in Hugo didn't give me a "Which wall do you mean, the north wall, the west wall, the northwest wall..." (etc) response. I'm still not sure if that is an Inform bug or library glitch or what.

I haven't used TADS 3, though, so I can't really comment on how things work with that language.

I wonder if someone might do a brief comparison of Hugo's world model with that of T3 and Inform.