With #7 we are returning content descriptions in turtle format, so we can start to build more complex and self-describing perceptions e.g. by using tags like sees, hears
Sometimes we will want the description to link to something interactive, e.g. it may be providing us with a complex event to display "you are seeing that someone is robbing a shop! What do you do?"
My proposal is for something simpler to start with, but I think no less exciting: since a describer can see what properties might be missing on the agent perceiving it, and we want to provide a fun and easy way for player's to build the role-playing of their characters - why don't we write a Describer which, picking up on lacking information, suggests to the user how their character could be extended. The user, making a simple "yes" or "no" choice can then start to build up their character's background
Open-ended but for example: "Does your character like football?"
Dependent on #11 since the Describer needs to know who is perceiving the object being described
There is also a concern to resolve with this design:
Later it will become relevant that certain properties like this could contradict eachother. For example:
"Does your character like sports?" -> no
"Does your character like playing football?" -> yes
With well-designed ontologies we could probably resolve this with inference, but since we want ontologies to be extensible and content/world servers to be modular, can we assume that they won't contradict or duplicate eachother's properties?
With #7 we are returning content descriptions in turtle format, so we can start to build more complex and self-describing perceptions e.g. by using tags like
sees,hearsSometimes we will want the description to link to something interactive, e.g. it may be providing us with a complex event to display "you are seeing that someone is robbing a shop! What do you do?"
My proposal is for something simpler to start with, but I think no less exciting: since a describer can see what properties might be missing on the agent perceiving it, and we want to provide a fun and easy way for player's to build the role-playing of their characters - why don't we write a Describer which, picking up on lacking information, suggests to the user how their character could be extended. The user, making a simple "yes" or "no" choice can then start to build up their character's background
Open-ended but for example: "Does your character like football?"
Dependent on #11 since the Describer needs to know who is perceiving the object being described
There is also a concern to resolve with this design:
Later it will become relevant that certain properties like this could contradict eachother. For example:
With well-designed ontologies we could probably resolve this with inference, but since we want ontologies to be extensible and content/world servers to be modular, can we assume that they won't contradict or duplicate eachother's properties?