Crossfire Mailing List Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: CF: Long term experimental ideas

-----Original Message-----
From: Hwei Sheng TEOH
Sent: 9/15/99 1:17 PM
Subject: RE: CF: Long term experimental ideas


Hmm, what's this obsession with Greater Demons? ;-) I must say that I'm
not particularly impressed with them, seeing that two icestorms (not even
large icestorms) from a level 70 wizard is enough to kill a Greater Demon.
But anyway, this is off topic... :-)
-----Begin Response-----       (<-- is that better Peter? :P)
I think it has to do with the fact that the graphic for the Greater Demon
was shamelessly stolen from a commercial game released in 1994 called Dark
Legions.  It is, consequently, the best looking graphic in the entire game.
It is also, incidentally, a copyright violation.  But in any case, it's the
coolest LOOKING demon I've seen in many a year.  :)

Hmm, this sounds *really* interesting. Perhaps we could even have
"dungeon guilds" which is basically a way for a player to sign up as an
"employee" at a particular dungeon, and he'd have to carry out orders from
the dungeon wizard for a reward (for low-level characters, a livelihood such
as a constant supply of food, for high-level characters other types of
bonuses, artifacts, etc.). Eventually, the player could even succeed the
-----Begin Response-----
That's certainly one mechanism that could be implemented, though I don't
know that I'd call that organization a guild.  But that's quibbling.  If
you're going to go to the extent of paying players, I wouldn't make it a
prerequisite to be officially "employed" by the wizard of a particular
dungeon.  The possibility of a player playing as a Kobold without the
controlling wizard even noticing if he doesn't go looking for it strikes me
as entertaining.

Unfortunately, that particular stunt opens the door to cheating.  A player
could create a throwaway character of one of the "monster" species a given
wizard is known to employ, infiltrate his domain, learn all the secrets and
traps and find out what other monsters work there, then invade with a
powerful warrior character forearmed with a lot of knowledge.

A lot of details like that would have to be dealt with and worked out
individually if we decided to go down this road.
But I perceive the implementation of this to be extremely complicated...
now, having human players for the "dungeon crew" would probably work fine,
but with AI monsters, things get a little tough... you'll need an AI that is
flexible enough to handle the ever-changing dungeon: suppose your dungeon
has a kobold servant whom you assigned to carry some stuff from your HQ to
an upper dungeon level. To account for the possibility that some adventurer
might be wandering down your dungeons, you'd want some way of knowing
whether that kobold actually made it to where he's supposed to go. You'd
also want to know about it if an adventurer is seen at a dangerously deep
dungeon level, etc..

Now generalize this scenario to a full-fledged dungeon with all kinds of
workers, with many different tasks, and you have to re-educate them every
time you  change the structure of your dungeon (else they'd get lost on an
errand),  plus you'll have to deal with racial squabbles, etc.. Very
complex, intricate web of communications and dynamic AI programming, etc..
I'm sure everyone has seen how "dumb" an AI games like C&C clones and
dungeon keeper have... well, they are very powerful already, yet they
require constant supervision by a DK who can see the bird's eye view of the
entire dungeon just by scrolling with the mouse. The AI we need here is one
fool-proof enough to be manageable by a player who is limited by normal
gameplay rules, and who probably wants to go on some quest rather than spend
the rest of his existence maintaining a dungeon.
-----Begin Response-----
Actually, I wouldn't characterize either the C&C Clone games or the Dungeon
Keeper game as having "very powerful" AI's.  Descent comes closer to that
accolade than any of those, and if you have seen the latest and greatest
iteration of Descent, there's an AI to impress.  I think you're right, it's
a tough problem.  But I also think it's doable, especially given that we can
do iterations to our heart's content, with a bunch of programmers to look at
the results and critique.  Rather than trying to code a monolith, we can
work our way in that direction.  The process of coding the facilities to
"own" a dungeon is going to take a while, and the AI can be iterated right
along with it.

The first step is to make the monsters conform to the same line of sight
restrictions a human player has to contend with.  But this can't be done
until the mechanism's David has been proposing to replace some of the
kludges map makers have built are in place, since there are maps that depend
on the cheating line of sight monsters have right now.

Everything is interrelated, if I may point out the obvious, so nothing is
going to happen over night.  If we can get some clarity on short, medium,
and long term goals, and get them written down, everything that does happen
can be directed to achieve those goals.  Things like, are we really going to
add lots of RPG elements and neglect hack'n'slash enthusiasts or not, are we
going to aim for none or some or all of my proposals, are we going to
implement David's portal or not, etc etc.  If we are going to become more
persistent as I've suggested, David's portal becomes a simple decision.
Implement it, and tweak the details as we go along.  If we're going to
retain an almost total lack of persistence, David's portal is problematic,
as you may have noticed from that thread.

I think we should put together a list of goals, withOUT a timeline
associated with any of them.  Just specify a sequence of progression for
interrelated goals, and map it out, starting by specifying goals everybody
can agree are going to be big complicated affairs to achieve, and then
breaking things up into smaller and smaller goals until we have something
manageable.  "Let players 'own' a dungeon" is a top level, long term, big
complicated affair.  "Enhance the NPC AI" is a subgoal of that.  "Fix NPC AI
line of sight behavior" is a subgoal of that.  "Enhance NPC dialog" is a
subgoal of enhancing the AI as well.  I think if we approach it like that,
our efforts at creating Artificial Stupidity will be closer to AI than AS,
and we CAN create an AI powerful enough to handle the complexities of
changing dungeon structures.  Sure it's complicated.  But we can do it.

[you can put yourself on the announcement list only or unsubscribe altogether
by sending an email stating your wishes to]