Crossfire Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Monster AI for version 1.0 CF (Was: Re: CF: Moving on to implement diseases)
- To: (Brian Thomas)
- Subject: Re: Monster AI for version 1.0 CF (Was: Re: CF: Moving on to implement diseases)
- From: (Philip Brown)
- Date: Tue, 3 Mar 1998 14:43:49 -0800 (PST)
- Cc: crossfire (at) ifi.uio.no (Crossfire Mailing List)
- In-Reply-To: <> from Brian Thomas at "Mar 3, 98 04:44:31 pm"
- Sender:
>>>>[From Brian Thomas]
> Scott writes:
>
> I do not believe the issue is developing a monster AI, but developing
> a monster (and possibly general object) scripting capability. Currently,
> monsters are controlled by some routines in CF with a minimal set of
> specified monster behaivors. This is not a great approach since any
> change in monster behaivor affects most all monsters and thus an improvement
> for one monster can make others do silly stuff.
>
I agree that the issue is not to hardcode all of
the monster behavior, I suppose I use the term "AI"
rather loosely (what I mean-- AI==monster behavior).
This is where a java server will kick butt.
All you have to do is define certain "interfaces" to objects such as
players, map, etc.
Then a monster type can have its own compiled "mymonster_AI.class", and the
server loads it in when appropriate. You can then get as strange as you
want, and you don't have to worry about making your own scripting language.
[to unsubscribe etc., send mail to ]
References: