Toradol
04-12-2001, 12:32 AM
Hello all. I've recently finished reading the SB article archive and having a quick tour of Castle Marrach. The articles, btw, are excellent and the technology is well thought out (reminds me a little of some of my MUD design pipe dreams in the early 90s).
What I wanted to ask about, however, was something I noticed in Castle Marrach that reminded me instantly and horribly of old MUD escapades... the problem of dealing with item ambiguity.
Item ambiguity arises when there are multiple objects that share a similar (or identical) description in the same area, and the player needs to specify one of them. ie:
>look
You are in a plain chamber. There is a tall courtier, a tall courtier and a fat courtier here.
>look at courtier
Which do you mean? The tall courtier, the tall courtier, or the fat courtier.
If you wanted to talk to one of the tall courtiers, and they have no unique adjective between them (something that can easily happen with 'generic' objects or NPCs) you're usually out of luck. Sometimes, (and I didn't test this fully in CM) you can get away with things like "look at the first tall courtier" but this can feel awkward.
So why doesn't ambiguity resolution utilize the prox system to help guess which object you are attempting to interact with? If in the above example one tall courtier was sitting on a bench, and I went and sat on the bench, then typed 'look courtier' or something similar, it would make sense to assume that I'm dealing with the closest courtier to me (if several are equally close we again devolve into prompting for an adjective or adjectives, but it seems that this prompting would become much less frequent).
As another example I occassionally want to check on myself or my belongings, but if I type "look at shirt" or "examine hands" in a crowded room it doesn't default to my shirt or hands, or even those of the people close to me, but asks which of the four dozen shirts in the room I'm interested in. Yes, the insertion of 'my' solves this reasonably simply, but what if I was trying to look at my friend's shirt, which turns out to be one of 5 brown linen shirts in the room. Figuring out which is his and then specifying it pulls me out of the game into arguing with the interface.
It seems like the prox system might offer some built-in help to this age-old problem.
Another solution, but one that would involve new coding, would be to retain a 'context' for an actor of the last object interacted with. So if I look at Sebastian, then look at shirt, it assumes I'm looking at Sebastian's shirt and not someone else's...
The system here has amazing potential to allow the creation of new and interesting online environments. I salute all involved.
If comments/suggestions of this nature are better suited to another forum please accept my apology. I did look before I lept...
What I wanted to ask about, however, was something I noticed in Castle Marrach that reminded me instantly and horribly of old MUD escapades... the problem of dealing with item ambiguity.
Item ambiguity arises when there are multiple objects that share a similar (or identical) description in the same area, and the player needs to specify one of them. ie:
>look
You are in a plain chamber. There is a tall courtier, a tall courtier and a fat courtier here.
>look at courtier
Which do you mean? The tall courtier, the tall courtier, or the fat courtier.
If you wanted to talk to one of the tall courtiers, and they have no unique adjective between them (something that can easily happen with 'generic' objects or NPCs) you're usually out of luck. Sometimes, (and I didn't test this fully in CM) you can get away with things like "look at the first tall courtier" but this can feel awkward.
So why doesn't ambiguity resolution utilize the prox system to help guess which object you are attempting to interact with? If in the above example one tall courtier was sitting on a bench, and I went and sat on the bench, then typed 'look courtier' or something similar, it would make sense to assume that I'm dealing with the closest courtier to me (if several are equally close we again devolve into prompting for an adjective or adjectives, but it seems that this prompting would become much less frequent).
As another example I occassionally want to check on myself or my belongings, but if I type "look at shirt" or "examine hands" in a crowded room it doesn't default to my shirt or hands, or even those of the people close to me, but asks which of the four dozen shirts in the room I'm interested in. Yes, the insertion of 'my' solves this reasonably simply, but what if I was trying to look at my friend's shirt, which turns out to be one of 5 brown linen shirts in the room. Figuring out which is his and then specifying it pulls me out of the game into arguing with the interface.
It seems like the prox system might offer some built-in help to this age-old problem.
Another solution, but one that would involve new coding, would be to retain a 'context' for an actor of the last object interacted with. So if I look at Sebastian, then look at shirt, it assumes I'm looking at Sebastian's shirt and not someone else's...
The system here has amazing potential to allow the creation of new and interesting online environments. I salute all involved.
If comments/suggestions of this nature are better suited to another forum please accept my apology. I did look before I lept...