Race Reference - Revision history

Fullwall at 00:40, 14 July 2013

2013-07-14T00:40:05Z

← Older revision		Revision as of 00:40, 14 July 2013
Line 20:		Line 20:
	behavior: # optional, a behavior root node (explained later)		behavior: # optional, a behavior root node (explained later)
	dialog:		dialog:
	allowed-rules: rule1, filename1 # comma-separated list of rules to allow (can be per-rule or per-file)		allowed-rules: rule1, filename1 # comma-separated list of rules to allow (can be per-rule or per-file). Leave empty for all rules allowed.

	The behavior root nodes are important for tribes. They are defined as a ''decision tree'' of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.		The behavior root nodes are important for tribes. They are defined as a ''decision tree'' of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.

Fullwall at 00:39, 14 July 2013

2013-07-14T00:39:43Z

← Older revision		Revision as of 00:39, 14 July 2013
Line 6:		Line 6:

	== YAML Info File Structure ==		== YAML Info File Structure ==
	Example YAML file:		Example YAML file showing all currently available keys:
	name: Name # required		name: Name # required
	tribe:		tribe:
Line 19:		Line 19:
	0: CREEPER		0: CREEPER
	behavior: # optional, a behavior root node (explained later)		behavior: # optional, a behavior root node (explained later)
			dialog:
			allowed-rules: rule1, filename1 # comma-separated list of rules to allow (can be per-rule or per-file)

	The behavior root nodes are important for tribes. They are defined as a ''decision tree'' of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.		The behavior root nodes are important for tribes. They are defined as a ''decision tree'' of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.

Fullwall at 00:34, 14 July 2013

2013-07-14T00:34:14Z

← Older revision		Revision as of 00:34, 14 July 2013
Line 36:		Line 36:
	selector -r: # retries children if they fail until a successful behavior completes		selector -r: # retries children if they fail until a successful behavior completes
	selector -p: # prioritises children		selector -p: # prioritises children
	js filename: "" # loads a javascript goal at the file location under your race folder and invokes the getBehavior(DataKey) function on it to get a Behavior instance.		js filename: "" # loads a javascript goal at the file location under your race folder and invokes the getBehavior(Tribe, DataKey) function on it to get a Behavior instance.
	your.package.name.ClassName: # loads the class at the given package and instantiates it to get a Behavior instance. Must have a 0-argument constructor (note that @Persist annotations from CitizensAPI will work for loading extra data).		your.package.name.ClassName: # loads the class at the given package and instantiates it to get a Behavior instance. Must have a 0-argument constructor (note that @Persist annotations from CitizensAPI will work for loading extra data).

Fullwall at 00:46, 13 July 2013

2013-07-13T00:46:35Z

← Older revision		Revision as of 00:46, 13 July 2013
Line 36:		Line 36:
	selector -r: # retries children if they fail until a successful behavior completes		selector -r: # retries children if they fail until a successful behavior completes
	selector -p: # prioritises children		selector -p: # prioritises children
	js filename: "" # loads a javascript goal at the file location under your race folder and invokes the getBehavior() function on it to get a Behavior instance.		js filename: "" # loads a javascript goal at the file location under your race folder and invokes the getBehavior(DataKey) function on it to get a Behavior instance.
	your.package.name.ClassName: # loads the class at the given package and instantiates it to get a Behavior instance. Must have a 0-argument constructor (note that @Persist annotations from CitizensAPI will work for loading extra data).		your.package.name.ClassName: # loads the class at the given package and instantiates it to get a Behavior instance. Must have a 0-argument constructor (note that @Persist annotations from CitizensAPI will work for loading extra data).

Fullwall at 00:41, 13 July 2013

2013-07-13T00:41:41Z

← Older revision		Revision as of 00:41, 13 July 2013
Line 28:		Line 28:
	We define this in YAML like so:		We define this in YAML like so:
	behavior:		behavior:
			sequence: # builds a sequence
			sub: # omitted
			sequence -r: # builds a sequence which retries children if they fail until the sequence is complete
	selector: # builds a random selector		selector: # builds a random selector
	sub: # omitted		sub: # omitted

Fullwall at 00:47, 11 July 2013

2013-07-11T00:47:44Z

← Older revision		Revision as of 00:47, 11 July 2013
Line 33:		Line 33:
	selector -r: # retries children if they fail until a successful behavior completes		selector -r: # retries children if they fail until a successful behavior completes
	selector -p: # prioritises children		selector -p: # prioritises children
	js filename: '' # loads a javascript goal at the file location under your race folder and invokes the getBehavior() function on it to get a Behavior instance.		js filename: "" # loads a javascript goal at the file location under your race folder and invokes the getBehavior() function on it to get a Behavior instance.
	your.package.name.ClassName: # loads the class at the given package and instantiates it to get a Behavior instance. Must have a 0-argument constructor (note that @Persist annotations from CitizensAPI will work for loading extra data).		your.package.name.ClassName: # loads the class at the given package and instantiates it to get a Behavior instance. Must have a 0-argument constructor (note that @Persist annotations from CitizensAPI will work for loading extra data).

Fullwall at 00:45, 11 July 2013

2013-07-11T00:45:46Z

← Older revision		Revision as of 00:45, 11 July 2013
Line 20:		Line 20:
	behavior: # optional, a behavior root node (explained later)		behavior: # optional, a behavior root node (explained later)

	The behavior root nodes are important for tribes. They are defined as a ~~*~~decision tree~~*~~ of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.		The behavior root nodes are important for tribes. They are defined as a ''decision tree'' of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.

	There are three flow constructs that help with behavior flows: sequences, selectors, and decorators. A sequence simply executes each behavior in a list one-by-one until they are all finished. A selector picks one or more behaviors from a list based on some kind of strategy, like a random selection or a priority-based selection. A decorator ~~*~~wraps~~*~~ another behavior without knowing what it does, for things like stopping it after a certain amount of time, or inverting its responses, or doing something before/after it completes.		There are three flow constructs that help with behavior flows: sequences, selectors, and decorators. A sequence simply executes each behavior in a list one-by-one until they are all finished. A selector picks one or more behaviors from a list based on some kind of strategy, like a random selection or a priority-based selection. A decorator ''wraps'' another behavior without knowing what it does, for things like stopping it after a certain amount of time, or inverting its responses, or doing something before/after it completes.

	Finally we have the edge nodes, which are user-defined and do actual work with the NPC. This is a very powerful construct that allows complex behaviors to emerge.		Finally we have the edge nodes, which are user-defined and do actual work with the NPC. This is a very powerful construct that allows complex behaviors to emerge.

Fullwall at 00:45, 11 July 2013

2013-07-11T00:45:24Z

← Older revision		Revision as of 00:45, 11 July 2013
Line 22:		Line 22:
	The behavior root nodes are important for tribes. They are defined as a *decision tree* of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.		The behavior root nodes are important for tribes. They are defined as a *decision tree* of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.

	There are three flow constructs that help with behavior flows: sequences, selectors, and decorators. A sequence simply executes each behavior in a list one-by-one until they are all finished. A selector picks one or more behaviors from a list based on some kind of strategy, like a random selection or a priority-based selection. A decorator wraps another behavior without knowing what it does, for things like stopping it after a certain amount of time, or inverting its responses, or doing something before/after it completes.		There are three flow constructs that help with behavior flows: sequences, selectors, and decorators. A sequence simply executes each behavior in a list one-by-one until they are all finished. A selector picks one or more behaviors from a list based on some kind of strategy, like a random selection or a priority-based selection. A decorator *wraps* another behavior without knowing what it does, for things like stopping it after a certain amount of time, or inverting its responses, or doing something before/after it completes.

	Finally we have the edge nodes, which are user-defined and do actual work with the NPC. This is a very powerful construct that allows complex behaviors to emerge.		Finally we have the edge nodes, which are user-defined and do actual work with the NPC. This is a very powerful construct that allows complex behaviors to emerge.

Fullwall at 00:45, 11 July 2013

2013-07-11T00:45:12Z

← Older revision		Revision as of 00:45, 11 July 2013
Line 20:		Line 20:
	behavior: # optional, a behavior root node (explained later)		behavior: # optional, a behavior root node (explained later)

	The behavior root nodes are important for tribes. They are defined as a decision tree of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.		The behavior root nodes are important for tribes. They are defined as a *decision tree* of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.

	There are three flow constructs that help with behavior flows: sequences, selectors, and decorators. A sequence simply executes each behavior in a list one-by-one until they are all finished. A selector picks one or more behaviors from a list based on some kind of strategy, like a random selection or a priority-based selection. A decorator wraps another behavior without knowing what it does, for things like stopping it after a certain amount of time, or inverting its responses, or doing something before/after it completes.		There are three flow constructs that help with behavior flows: sequences, selectors, and decorators. A sequence simply executes each behavior in a list one-by-one until they are all finished. A selector picks one or more behaviors from a list based on some kind of strategy, like a random selection or a priority-based selection. A decorator wraps another behavior without knowing what it does, for things like stopping it after a certain amount of time, or inverting its responses, or doing something before/after it completes.

Fullwall: Created page with "== Creating a Race == Race folders need to go in [server folder]/plugins/Adventures/races/YourRaceHere. Every race needs a descriptor, or a file that tells Adventures what to ..."

2013-07-11T00:43:17Z

Created page with "== Creating a Race == Race folders need to go in [server folder]/plugins/Adventures/races/YourRaceHere. Every race needs a descriptor, or a file that tells Adventures what to ..."

New page

== Creating a Race ==
Race folders need to go in [server folder]/plugins/Adventures/races/YourRaceHere.
Every race needs a descriptor, or a file that tells Adventures what to load for your race. This will be info.yml OR info.js OR info.extension, where extension is an embedded Java script language such as Jython or JRuby.

If a scripting language is used for the info file, it will need to have a getDescriptor() method in the root namespace which returns an instance of net.citizensnpcs.adventures.race.RaceDescriptor.

== YAML Info File Structure ==
Example YAML file:
name: Name # required
tribe:
max-size: 10 # required
behavior: # optional, a behavior root node (explained later)
members:
names: # required, a list of possible names for new members to have, picked at random
0: Bob
1: SirBob
2: Bob2
types: # required, a list of possible entity types for new members to have, picked at random
0: CREEPER
behavior: # optional, a behavior root node (explained later)

The behavior root nodes are important for tribes. They are defined as a **decision tree** of behaviors, which are a single unit of work for an NPC. A tree is a way of organising data which looks similar to an upside-down tree. For a simple illustration, look [https://monk.library.illinois.edu/cic/public/analytics/decision_tree.png here]. The NPC's AI will start from the top and work its way down to a behavior.

There are three flow constructs that help with behavior flows: sequences, selectors, and decorators. A sequence simply executes each behavior in a list one-by-one until they are all finished. A selector picks one or more behaviors from a list based on some kind of strategy, like a random selection or a priority-based selection. A decorator **wraps** another behavior without knowing what it does, for things like stopping it after a certain amount of time, or inverting its responses, or doing something before/after it completes.

Finally we have the edge nodes, which are user-defined and do actual work with the NPC. This is a very powerful construct that allows complex behaviors to emerge.

We define this in YAML like so:
behavior:
selector: # builds a random selector
sub: # omitted
sub2: # omitted
selector -r: # retries children if they fail until a successful behavior completes
selector -p: # prioritises children
js filename: '' # loads a javascript goal at the file location under your race folder and invokes the getBehavior() function on it to get a Behavior instance.
your.package.name.ClassName: # loads the class at the given package and instantiates it to get a Behavior instance. Must have a 0-argument constructor (note that @Persist annotations from CitizensAPI will work for loading extra data).