Travel Connectors and Barriers
TravelConnectors
In addition to Doors, the direction property of a Room can point to a TravelConnector (a Door is in fact a particular kind of TravelConnector). A TravelConnector is a kind of object that allows you to define:
- Whether or not the exit should be listed in the exit lister.
- Whether or not the exit is apparent to the player character (if not, then obviously it cannot be used).
- Whether or not travel should be allowed in that direction according to any condition you care to define.
- Any side-effects of travel, such as displaying a message describing the travel.
- Whether or not the player character knows where the exit leads (this is relevant to the routeFinder).
The methods and properties that enable a TravelConnector to be used for these purposes are as follows:
- destination (a Room) Where this TravelConnector (and hence the exit to which it is attached) leads to. This would normally be set to a Room.
- isConnectorApparent (true or nil) This must be true if travel is to be allowed via this connector. By default this property is true. In this context 'apparent' means 'not concealed'. To be usable the TravelConnector must be not only apparent but visible (i.e. isConnectorVisible must be true); this is the case if isConnectorApparent is true and the lighting conditions are sufficient to see the exit by.
- isConnectorListed (true or nil) If this is true then the corresponding exit will be shown in the exit lister. This can be set to nil to hide the exit (presumably until something occurs to reveal it, when you would then set it to true). By default isConnectorListed takes its value from isConnectorApparent.
- isDestinationKnown (true or nil) This is true when the player character knows where this exit leads. By default this is true if the destination of the connector has been visited (visited = true). You can, however, define it as true at the start of the game if the player character is meant to start out knowing where the exit leads (though it may be easier to define familiar = true on a region than to set isDestination = true on a whole lot of TravelConnectors). This property is mostly relevant to the routeFinder (the mechanism behind the GO TO command). Note that a Room is a TravelConnector whose destination is itself (so that if an exit leads directly to a room the destination of that exit is known if the destination room has been visited). Note also that a slightly different rule applies to doors: the isDestinationKnown property of both sides of a door becomes true once the player character passes through one or other side of it.
- noteTraversal(actor) This method is called when actor traverses this connector, and can be used to carry out any side-effects of the travel. By default we simply display the travelDesc here if actor is the player character.
- travelDesc (double-quoted string) A message that's displayed when the player character traverses this TravelConnector (alternatively, this could be defined as a method that carried out all sorts of side effects of traversing this TravelConnector). Note that if noteTraversal() is overridden without a call to inherited(), the travelDesc will be disabled. Note also that the default behaviour of travelDesc is to call the TravelConnector's doScript() method if the TravelConnector is also a Script; this makes it easy to vary the message displayed when the player character traverses the connector by adding an EventList class to the TravelConnector's class list and defining an eventList property on it.
- execTravel(actor, traveler, conn) This method is called by travelVia(actor) to move the actor to the destination when travelling via the corresponding exit. You won't normally want to override this method unless you want it to do something drastically different like killing the player character or ending the game for some other reason. The normal behaviour of this method is defer to the execTravel() method of the room being travelled to, which actually carries out the travel and issues all the appropriate travel notifications: in particular it issues the before travel notifications (by calling beforeTravel(actor, self) on every object in scope, including the current location), then, if the actor is the player character it displays the travelDesc message (if any) and notes that the player character now knows where this TravelConnector leads. It next calls execTravel(actor) and finally carries out the after travel notifications on the new location (by calling afterTravel(actor, self) on every object in scope in the new location).
- travelVia(actor) This is the principal method called when the actor tries to traverse this TravelConnector. It first checks for travel barriers (see below), and then if the travel barriers don't prevent travel, it calls the execTravel() method. The only reason to override travelVia() would be if you wanted attempting to travel via this connector to do something drastically different from a normal travel action.
- canTravelerPass(traveler) Returns true if the traveler is allowed to pass through this TravelConnector. Return nil to prevent travel (normally travel would be blocked conditionally, e.g., perhaps only certain objects would be allowed to pass, or the player character is only allowed to pass if he's wearing the magic ring or not carrying the great big ladder that won't fit through the narrow gap). The traveler parameter would normally be the traveling actor, but might be a vehicle in which the actor is traveling.
- explainTravelBarrier(traveler) If canTravelerPass(traveler) returns nil (or might return nil), explainTravelBarrier(traveler) should display a message explaining why travel is disallowed.
- travelBarriers An optional list of TravelBarrier objects (or a single TravelBarrier object) that enforce further conditions on traveling via this connector (see further below).
- transmitsLight: can be used to control whether or not a connector (typically a door) that leads to a lit room is visible from a dark room. This can be used, for example, to determine whether or not the player character can use a door from a dark room to a lit one. This is true by default.
Note that the parameter of canTravelerPass() etc. is called traveler: it will normally be the actor doing the traveling but it could also be an object that an actor is attempting to push through this TravelConnector (via a command like PUSH TROLLEY NORTH, for example). This means that canTravelerPass(traveler) can also be used to selectively prevent the pushing of objects in certain directions; for example, you might want to use this method to prevent a heavy object being pushed up a flight of stairs.
An example of a TravelConnector which conditionally blocks travel could be one that only allows travel down a smoke-filled passage if the player is wearing a wet blanket, which could be defined thus:
landing: Room 'Landing' 'landing' "The smoke is already becoming so thick here that it's hard to see much. Your bedroom lies to the north -- if you can make your way through the smoke. Most of the other upstairs rooms are down the passage the other way, to the south, but the worst of the smoke seems to be coming from there. " down = landingStairs north: TravelConnector { destination = bedroom travelDesc = "You manage to force your way through the smoke, coughing and choking as you go. " canTravelerPass(actor) { return blanket.wornBy == actor && blanket.isWet; } explainTravelBarrier(actor) { if(blanket.wornBy == actor) "You take a few steps down the corridor but the smoke forces you back as the blanket starts to get singed. "; else "The smoke is too thick; you find yourself coughing and choking after the first step and are forced to retreat. "; } } south { "The smoke is too thick that way; you almost choke to death with the first step south you take. Well, it's not as if there's anything down there you really need all that much right now. "; } regions = upstairs ;
This incidentally illustrates that if we only want to use a particular TravelConnector once, we don't need to define it as separate named object, since we can instead define it as an anonymous nested object directly on the appropriate direction property.
Note that a TravelConnector only establishes a connection in one direction: the TravelConnector defined on the north property in the example above creates a direction north from the landing to the bedroom, but no connection back south from the bedroom to the landing. Often, as here, that's what we want (since we wouldn't want the same travel restrictions to apply to the attempt to return from the bedroom to the landing), but if you do want a TravelConnector that works the same both ways, you could try using the SymConnector defined in the symconn extension.
TravelBarriers
For many simple cases conditionally preventing travel by using the canTravelerPass() method of a TravelConnector will suffice, but there may be situations where it's not the most convenient way to do it. The two most common situations where this might arise are where:
- You want to apply the same condition (and display the same refusal message if it's not met) to a number of TravelConnectors, and it would be tedious to have to define the same canTravelerPass() and explainTravelBarrier() methods on each of them.
- Several different conditions apply on a TravelConnector, each requiring a different explanation if it's not met, so that the explainTravelBarrier() method would need to contain a messy mass of if statements that largely duplicate those in the canTravelerPass() method.
If either or both of those two conditions obtain, you might be better off defining and using a TravelBarrier object to represent the conditional barrier to travel. There are basically two methods you need to define when creating a TravelBarrier object:
- canTravelerPass(traveler, connector) Return true if traveler is allowed to traverse connector (the TravelConnector object to which this TravelBarrier instance is attached) and nil otherwise.
- explainTravelBarrier(traveler, connector) Explain why the traveler is not permitted to traverse this connector if travel is disallowed.
Note the additional connector parameter on these methods. This is most likely to be useful when you want the explainTravelBarrier() method to name the connector that's being blocked. For example, suppose our game had several exits that required the player character to be wearing a wet blanket in order to penetrate the smoke. We might set it up thus:
landing: Room 'Landing' 'landing' "The smoke is already becoming so thick here that it's hard to see much. Your bedroom lies to the north -- if you can make your way through the smoke. Most of the other upstairs rooms are down the passage the other way, to the south, but the worst of the smoke seems to be coming from there. " down = landingStairs north: TravelConnector { destination = bedroom travelDesc = "You manage to force your way through the smoke, coughing and choking as you go. " theName = 'the corridor' travelBarriers = [smokeBarrier] } south: TravelConnector { destination = bathroom travelDesc = "With the aid of the blanket, you are able to make your way through the smoke to the bathroom. " theName = 'bathroom passage' travelBarriers = [smokeBarrier] } regions = upstairs ; smokeBarrier: TravelBarrier canTravelerPass(actor, connector) { return blanket.wornBy == actor && blanket.isWet; } explainTravelBarrier(actor, connector) { if(blanket.wornBy == actor) "You take a few steps down <<connector.theName>> but the smoke forces you back as the blanket starts to get singed. "; else "The smoke is too thick; you find yourself coughing and choking after the first step and are forced to retreat. "; } ;
Note how we've added a custom theName property to the two TravelConnectors so that smokeBarrier.explainTravelBarrier(actor, connector) can refer to them.