It's described in Game Programming Gems 7 about bot logic for Unreal
Tournament and sounds pretty exciting, it's not Unreal Script but
other component (don't remember name now I have no access to this book
atm)

2008/7/14 Ross Brackstone <[EMAIL PROTECTED]>:
> Hm Fuzzy state system is a weird description to me, you can push them onto a
> stack in UE3, but from the UDN documentation:
>
> In UnrealScript, each actor in the world is always in one and only one
> state.
>
> http://udn.epicgames.com/Three/UnrealScriptReference.html#States
>
> On Mon, Jul 14, 2008 at 8:43 PM, Dariusz Dawidowski
> <[EMAIL PROTECTED]> wrote:
>>
>> Unreal3 uses state machine like you said but even better - every state
>> has weight so it's Fuzzy State Machine.
>> But do we have manpower to rewrite quest system ? I doubt.
>>
>> Cheers
>>
>> 2008/7/13 Amir Taaki <[EMAIL PROTECTED]>:
>> > Hi!
>> >
>> > Before there was talk about possible structures for FSM's in CEL during
>> > cspop
>> > for managing AI but this proposal was ousted after it was seen that
>> > quests
>> > could be applicable in this sense.
>> >
>> > Now with making the movement controller for apricot character the
>> > various
>> > movesets have to be made and coordinated together. A monolithic design
>> > in the
>> > first sense may seem simpler and work better but it quickly becomes
>> > apparent
>> > in it's shortcomings. Namely,
>> >
>> >  1. Methods clutter whereby functions start getting grouping prefixes to
>> > distinguish themselves.
>> >  2. Extensibility by being able to tack on new components.
>> >  3. Configuration- monolithic is essentially saying "Here's what you
>> > get"
>> >
>> > A micro component design by contrast offers clarity at the expense of
>> > higher
>> > complexity and lower ease of use. The complexity comes from coordinating
>> > the
>> > various micro components to work together in a coherent way.
>> >
>> > Typically when you perform an action in-game, there will be a few
>> > actions you
>> > cannot perform while that action is being performed. Examples can be
>> > that you
>> > can jump while spin-attacking but cannot grab ledges, or that you cannot
>> > jump
>> > while performing a roll.
>> >
>> > Maybe pressing jump during a roll while wait for the roll to finish then
>> > perform an extra long rolling-jump. Notification of the roll being
>> > finished
>> > is not a problem in CEL as there is already a mechanism (messages). But
>> > for
>> > seeing that a roll is being performed the current best solution is,
>> >
>> >  csRef<iPcRoll> roll = celQueryPropertyClassEntity<iPcRoll> (entity);
>> >  if (roll->IsRolling ())
>> >    // blaa
>> >
>> > Not a problem, except due to the tight coupling with iPcRoll which
>> > totally
>> > destroys any reason for moveset components.
>> >
>> > Another issue is knowing what action to perform depending on a state
>> > dependent
>> > in another component-
>> >
>> >  def JumpPressed (self, entity):
>> >      if self.jump.IsJumping ():
>> >          self.jump.Jump ()
>> >      else
>> >          self.glide.Glide ()
>> >
>> > Something more convenient and transparent for the user could be
>> >
>> >  def __init__(self, entity):
>> >      self.BindAction (JumpPressed, (self.Glide, self.Jump))
>> >
>> > With the glide knowing beforehand automatically which state it is
>> > allowed to
>> > transition from and also with the option to have the user to override
>> > this
>> > and specify themselves how they want.
>> >
>> > Essentially I'm asking what mechanism already exists in CEL (are quests
>> > sufficient) to have a state machine easily accessible by multiple
>> > components
>> > for various actions like,
>> >
>> >  1. Can register more than one state- object can be in multiple states
>> > which
>> > can be removed individually. Like attacking, jumping and analogmotion.
>> >  2. Ruleset to describe what action to execute in what components
>> > depending
>> > on existing states.
>> >      Component can then ask permission to register a state- if the
>> > ruleset
>> > rejects the new state then the action doesn't get performed.
>> >
>> > Thanks for your time,
>> >
>> > Genjix
>> >
>> >
>> > -------------------------------------------------------------------------
>> > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
>> > Studies have shown that voting for your favorite open source project,
>> > along with a healthy diet, reduces your potential for chronic lameness
>> > and boredom. Vote Now at http://www.sourceforge.net/community/cca08
>> > _______________________________________________
>> > Cel-main mailing list
>> > Cel-main@lists.sourceforge.net
>> > https://lists.sourceforge.net/lists/listinfo/cel-main
>> >
>>
>> -------------------------------------------------------------------------
>> Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
>> Studies have shown that voting for your favorite open source project,
>> along with a healthy diet, reduces your potential for chronic lameness
>> and boredom. Vote Now at http://www.sourceforge.net/community/cca08
>> _______________________________________________
>> Cel-main mailing list
>> Cel-main@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/cel-main
>>
>
>
>
> --
> ~ Ross Brackstone
> -------------------------------------------------------------------------
> Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
> Studies have shown that voting for your favorite open source project,
> along with a healthy diet, reduces your potential for chronic lameness
> and boredom. Vote Now at http://www.sourceforge.net/community/cca08
> _______________________________________________
> Cel-main mailing list
> Cel-main@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/cel-main
>
>

-------------------------------------------------------------------------
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
_______________________________________________
Cel-main mailing list
Cel-main@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/cel-main

Reply via email to