Chloroform And Blackjacks

[FishFace]

I'm not absolutely sure they didn't just use standard heights for their mantle things...

 

I see what you mean about the problems with mantling, now - you needn't implement jump-mantles; I think that should remain the sole department of the Thief, so you can escape, and because of the difficulty involved.

[bob_arctor]

TDS had simplified ladders. It is possible. In fact preferable to broken TII style ladders: LeT ME OFF... Here Goes /jump CLANK CLANK HEY GET HIM!!!!

Or Climb to top and get off.... arrgGGGGggggGGG I'm falling down the mineshaft!!

[bob_arctor]

Ultimately if you were a guard valuing your life you wouldn't follow a mentalist (ie exploit using bow-mad knifer) thief up a ladder while they were there. Hence having them climbing ladders out of sight instantly, (cheat, natch) would be more sensible. When the player is there they could run and get an archer/musketeer (?....Please?) and start shooting.

[Maximius]

On the topic of chloroform, heres an idea:

 

The Thief has both a black jack and some kind of KO liquid like chloroform. He can conk a guard with the jack, but the guard will wake up in like 10 minutes unless he also uses the chloroform on him which puts him down for the game.

 

To avoid the problem of modelling a guard waking up and getting to his/her feet, the games objectives could simply be set so that the player fails the mission if a BJed guard is allowed to awaken. You never actually have to see the guard awaken, because if he does the game is over. This seems to me to be reflective of real life, if a guard wakes up after being cracked over the head, the hold castle is going to go nuts looking for the perpetrator of the crime.

 

So here is how it works. I blakjak a guard, but then I have only about 10 minutes to drag him to a dark corner and apply the chloroform. If I do not, he will awaken and my mission is failed because now the guards will go bananas looking for me, making completing the mission impossible.

 

I think this is valuable for two reasons: One it makes the blackjacking of guards a more complex proposition, you have to hide them and you have to take the time to chloro them and you only have about 10 minutes to do so before they wake up. I can forsee some interesting situations where you KO a guard with the jack, but then have to run and hide from someone else, and the clock is ticking while you wait to run back to the room and chloro his ass so he doesnt wake up and raise hell.

 

Two it allows for the IDEA of guards awakening after being hit with the blackjack and having a real impact on the game, w/o actually having to model them waking up with all the attendant work and headaches.

 

comments?

[Springheel]

To avoid the problem of modelling a guard waking up and getting to his/her feet, the games objectives could simply be set so that the player fails the mission if a BJed guard is allowed to awaken.

 

So you KO a guard, walk away and ten minutes later the screen goes black and "mission fails" comes up? Not fun. And it is very arbitrary. Why should a guard waking up mean you fail? Should you immediately fail the game if you set off an alarm?

 

Ten minutes is more than enough time to hide the body and chloroform it, so it doesn't add any complexity unless you don't have enough chloroform. But the result of that system would be players simply killing guards to make sure they don't wake up. Then you have to impose a 'no kill' objective....

 

When you start having to add new restrictions to the game just to justify adding a new tool, it suggests to me the tool isn't worth having.

[oDDity]

No one likes the chloroform thing Maximus, no matter what the details. IT's just not fiting.

[oDDity]

]A real life guard on a ladder has many choices. If they are attacked from below they might hang by one hand and draw a weapon.  They might jump off the ladder.  They might kick their pursuer in the head or step on his hands.  They might swing around to the other side of the ladder, or slide down the rails quickly.  Can you imagine the amount of work to code and animate these options for AI?

] None of that is necessary it he game. all that is required is that the AI has an above attack and a below attack while on a ladder, so the player cannot stand at the top or bottom and hack the AI to pieces.

It's not equal because AI will never be anywhere near as smart as the player.  The player can lure several AI to a ladder, block the top and then attack the guards one at a time as they try to get off.  AI are not going to be able to do the same.  AI are not going to push a barrel in front of the ladder while the player is on it.   Etc.

By equal I mean that the same tactic is open to both player and AI. Obviously the player will always be better at exploiting it.

None of those things are exploits though. Pushing a barrel in front of a ladder is a legitimate tactic. If the guards have the two attacks while on the ladder, luring them there won't matter. If the player wants to have fun blocking ladders, then let him.

I suppose they shouldn't have included elevators in the Thief series because you could lure AI there and crush them? THat hardly amounts to a buggy game. It's just a fact od life that the player will always find ways to trick the AI no matter what you do.

Whe THief came out, it had some of the best AI ever seen in a game, and it's still great, but there is still a dozen different ways to exploit it. THere's no shame in it.

You can exploit the crap out of the AI in the Baldur's Gate series, and it'sstill by far my favourite game.

None of these things are reasons for not inculding AI ladder climbing.

Obviously nothing is stopping people from trying to implement ladder climbing once the mod is released, but I wouldn't be happy working for half a year or more to release a feature that doesn't work properly.

Define 'properly'

IF you're saying 'properly' means there is no way they player can gain an unequal advantage over the AI, then forget it, because the player will always find ways to do that in any part of the game, not just with ladders.

 

What about where the player is behind him, or in front at the top? Guards fighting on the level have several different animations for sword fighting, enabling them to do complex things such as parry and so on.

An above swing and a below swing will stop the player blocking at the top or bottem and using melee on the guard. THe below swing will be a circular slash low and behind the guard, and the above one will be similar. but above his head.

 

 

But totally unrealistic. It would ruin the variety, and part of tactics is knowing that _some_ places you should be able to access without being followed (e.g. via rope arrows) and ranged weapons for everyone would defeat this.

There's nothing unrealistic about all the guards carrying bows or crossbows - unless there's some sort of wood shortage in the game that I don't know about.

Ha, and now you want us to deliberately give the player an advantage by letting him climb safely about, knowing that some of the poor dumb AI don't even have ranged weapons to combat him with. THat's programming explotation into the game, since the player can easily exploit the non ranged guards with his bow and climbing ability.

THe player already has enough of an advantage, he doesn't need any extra.

[obscurus]

OK. I don't know how many of you have held or wielded a broadsword, but you are going to have a bit of trouble weilding one on a ladder - by the time you swing your prey will have moved a couple of rungs out of range. It is not impossible by any means, but it is foolish. The best attack when chasing someone up a ladder is to grab the leg of the thief and pull him out and thus throw him to the ground. The best tactic when being chased up a ladder is to stomp on your pursuer's fingers. Try it with some friends. I think that should be simple enough to animate. For the sake of simplicity, you can dispense with weapon attacks by the pursuing guard, since they are not the wisest option anyway (but if you like the idea of guards weilding swords while climbing ladders, I don't really think it too hard to animate, even if it is not very realistic).

 

Now, build your ladders like this:

 

1. create a ladder volume, a rectangular prism which runs the height of the ladder - this defines the dimensions of the ladder space.

2. place an entity at the top and bottom of the ladder - this defines the entry and exit points of the ladder (special pathnodes for the AI)

3. Place an entity at the bottom, and one at the top of the ladder, which marks a'hotzone' radius where if the player is standing, the AI will not climb the ladder, but will either pull out a ranged weapon, or run for help. If the player moves away from the hotzone, the AI will move towards the ladder base pathnode, and from there to the top pathnode. If the player moves into the hotzone while the AI is climbing, the AI will move back down the ladder.

 

I am not a coder but I cannot imagine that would require complex AI or a huge amount of coding - it is very simple. This type of method works beutifully in quite a few games. If you expect to get perfectly matched animations to varied geometry, forget it - you would need to have real time inverse kinematics to really do that. Just make ladders a standard width and size, and accept the limitations of the system. Mappers will need to be careful to make sure their ladders conform to the standard.

 

 

The whole point of the game is to exploit the AI's weaknesses to complete the mission. Messing around with the AI on a ladder is a perfectly reasonable exploit, as is throwing down barels and crates onto a guard from above, or sneaking past a guard silently. If you couldn't explot the weaknesses of the AI, you wouldn't have much of a game.

 

And if ladders are being used so much in a mission that it unbalances the game in favour of AI or thieves, depending on what you implement, then it is a case of flawed level design.

 

And it seems like people are saying "I can't program ladders well enough, so they won't be included" when they have an overly involved model in mind. You don't need to make it very complex to get a reasonably effective result. You don't need to consider every possible permutation of ladder climbing to get a decent result. It is a game - you can't model every aspect of reality perfectly, but you can use tricks to make it seem like you have, and the player won't notice the difference. Build a simple base model and refine it from there, don't try to have perfect ladders in one go.

 

Ask yourself this: What would I do if I were following a theif armed with a bow and arrows and he ran up a ladder? I would not chase him unless I was sure I could catch him at the bottom, and if he got away, I would try to make sure that the thief had no way down but into the waiting swords of several guards, or starving to death in a high place. I would not climb the ladder until I was sure I was out of range of any thiefy arrows.

 

If you would engage in combat on a ladder (other than trying to grab the legs of the guy above you), you are an idiot - both you and your opponent will probably fall off pretty quickly, and if you were high up, you would both most likely die. It takes enough energy to climb a tall ladder with 40+ kilos of armour, a 5 - 10 kilo sword, and being a bulky guard, your own body weight will present enough of a challenge without adding aerial combat into the mix. Take your sword out of it's scabbard and you now have one hand supporting at least 60 kg, and the other trying to wield a heavy broadsword, and your thief is gonna get away while you cling on for dear life out of exhaustion... If you try to attack someone from below with your sword while they are grabbing at your legs, you will probably fall as you have one hand and one foot on the ladder supporting a load of weight while someone is tugging at your boots.

 

I really suggest you guys get a ladder, a few mates, some padding, and roleplay some scenarious to see just how realistic certain ladder based activites are before you dismiss the concept of ladders in the game outright because you have an overly complicated model in mind. Seriously. It'll be fun.

Edited August 24, 2005 by obscurus

[FishFace]

By equal I mean that the same tactic is open to both player and AI. Obviously the player will always be better at exploiting it.

None of those things are exploits though. Pushing a barrel in front of a ladder is a legitimate tactic.

 

The mentioned exploit was pushing barrels to block a ladder. Now I know "not pushing barrels anywhere" could be construed as engaging the tactic really really badly, but no. The guards will not be equal if you are able to do this.

 

It is an exploit because you, the player know exactly what to do: jump off, or climb up, or somehow move/destroy the barrel. The guard cannot do this, therefore it is an exploitation of the AI's limits, not a tactic that should work. Luring guards beneath elevators is valid, BUT the guard should then bugger off as the elevator comes down. Otherwise it is still an exploit.

 

Whe THief came out, it had some of the best AI ever seen in a game, and it's still great, but there is still a dozen different ways to exploit it. THere's no shame in it.

 

There is shame: it's lazy and detracts from the game. What would you rather have - buggy AI or good AI? Thought so.

 

There's nothing unrealistic about all the guards carrying bows or crossbows

 

It's a completely useless way of getting around a limitation. Do FM authors want all their guards to have ranged weapons? Would every armed member of staff in every building and street in every setting in The City have a ranged weapon? Of course they wouldn't. While we're at, lets make everyone a guard - it's only slightly less realistic than what you propose.

 

 

 

@Obscurus: It still wouldn't work, as you could hold an AI on the ladder indefinitely by moving in and out of the 'hotzone.' Such predictability would look awful. It would certainly be easy to program, once you have the basic vertical pathfinding, but it would be rubbish!

The problem of the matter is that it is impossible to give guards anything like the ability of players to exploit their vulnerability. If guards could block the player in, or deal with the players attempts of ladder exploitation, then it would be fine, but it's not feasible. Exploiting AI weaknesses should be restricted to realistic weaknesses. I don't think anyone wants to be able to jump up and down until an AI blunders up to you, whereupon you smack them in the face with a blackjack. While the exploit (ambush/blockading) is valid, the response (getting KOd/getting stuck) is not.

 

If we can't have ladders that allow you to use them as FM authors will want (and could do previously) it would be awful. It's not so much problems of standard width - more the fact that the AI will have difficulty getting on and off if jumps or turns are required.

 

This "overcomplicated model" is what would be necessary. You'd have to program the guards for many things the player could validly do, otherwise their response would be pathetic and unrealistic.

 

Think: what would you do if you reached the top of a ladder and there was a crate in the way? You'd move it out of the way, perhaps throw it down. A guard would have to be programmed and animated to do this. Every single thing that a player could (in probability) bring against the guard should be programmed for. This is a can of worms.

[Maximius]

So you KO a guard, walk away and ten minutes later the screen goes black and "mission fails" comes up?  Not fun.  And it is very arbitrary.  Why should a guard waking up mean you fail? Should you immediately fail the game if you set off an alarm?

 

Ten minutes is more than enough time to hide the body and chloroform it, so it doesn't add any complexity unless you don't have enough chloroform.  But the result of that system would be players simply killing guards to make sure they don't wake up.  Then you have to impose a 'no kill' objective....

 

When you start having to add new restrictions to the game just to justify adding a new tool, it suggests to me the tool isn't worth having.

<{POST_SNAPBACK}>

 

RIght, so you dont just walk away, you chloro them to finish the job.

I prefer to play with the "no kills" objective anyway, so it would not be so very different than that. A waking guard *knows* he has been KOd, so the whole place would go nuts knowing there is a thief about. Which to me seems a compelling reason to fail a player. And yes, I think setting off an alarm should fail the mission, at least in the missions Id like to make.

 

The ten minutes was just an arbitrary time, it could be 10 seconds is a mission author wished. I just thought it would add an element of realism, black jacks arent going to KO someone for hours on end so you add the chloro to make it stick, but that brings its own risks, taking the time to do it and all plus the possiblity of running out of the stuff.

 

But all that being said, it was just an idea, if no one likes it, fine, I just go and sulk.

Hmmp.

 

 

No one likes the chloroform thing Maximus, no matter what the details. IT's just not fiting.

<{POST_SNAPBACK}>

 

 

Im sulking. See above.

[Domarius]

The ONLY question regarding AI climbing ladders is whether or not you can give the AI a reasonably realistic way of making the judgement call as to whether they should follow you up or not..

And that is the can of worms we do not want to open at this stage.

 

In real life if you chased someone onto a ladder, you would try and pull them down, coordinate an ambush at the top, poke at them from above, etc. etc. and if you were caught halfway down a ladder, you could jump off, swing around to the other side, slide down, jump to a nearby ledge, etc. etc.

 

...all these things are nessecary to make AI ladder climbing a reasonable choice for the AI (if they want to avoid being exploited - since you can do most of the above things yourself) and all of which are very very organic, freeform actions that would require heaps of new anims and programming.

[oDDity]

It is an exploit because you, the player know exactly what to do: jump off, or climb up, or somehow move/destroy the barrel. The guard cannot do this, therefore it is an exploitation of the AI's limits, not a tactic that should work. Luring guards beneath elevators is valid, BUT the guard should then bugger off as the elevator comes down. Otherwise it is still an exploit.

You could define any situation in which the AI doesn't behave as intelligently as the player to be an exploitable one. That's means the entire game is exploitable, of course.

On the one hand you're not happy with having an intellectual advantage over AI on ladders, and on the other you want to leave them open to exploit by leaving them at a huge disadvantage by having no range weapon. I think the AI need all the advantage they can get.

 

There is shame: it's lazy and detracts from the game. What would you rather have - buggy AI or good AI? Thought so.

THere's a difference between good AI and perfect AI. Perfect AI is impossible, and has never been created in a game. It's always open to exploit. I'm willing to trade some possible expolit by a small number of insignificant 'players', for a feature I want in the game.

IF you want to play your game by luring AI to ladders and killing them, that's your problem. YOu can't blame us for giving you the ability to do so. That's like stabbing someone to death and them blaming the person who sold you the knife. If you can't exercise enogh self control to not pick exploits in the AI so you can cheat yourself, then you shouldn't be playing games.

It's a completely useless way of getting around a limitation. Do FM authors want all their guards to have ranged weapons? Would every armed member of staff in every building and street in every setting in The City have a ranged weapon? Of course they wouldn't. While we're at, lets make everyone a guard - it's only slightly less realistic than what you propose.

It's not getting around any limitation. I always wanted all lhe guards in Thief to have both range and melee weapons, and I want the same for TDM.

Every AI that matters on a level is a guard. The rest are noncombatants who don't chase the thief anyway. The preists and mages have ranged spell attacks. And I'd like the undead to have the kind of attack the xombies in RtCW had, where they'd throw ghostly skulls at you which blacked out your screen, left you short of breath and drained your health. ATM, the xombies in Thief are a total waste of space. All combatant AI should have a range attack of some sort, so the player can't fuck them around by mantling or kiting them.

[Domarius]

All combatant AI should have a range attack of some sort, so the player can't fuck them around by mantling or kiting them.

Yep, we're already going to do this:

http://forums.thedarkmod.com/index.php?showt...indpost&p=11432

(That's a private board, someone can quote it if they want)

Perfect AI is impossible, and has never been created in a game. It's always open to exploit.

Yes you'll never get rid of all exploit's 100%, but you need to minimise it, and some of us feel this would be too big an exploit.

If you can't exercise enogh self control to not pick exploits in the AI so you can cheat yourself, then you shouldn't be playing games.

In my opinion, the goal of any game where you have an opponent is to look for weaknesses in your opponent and exploit those weaknesses. In a real life sport game, the rules of what you can and cannot do are clearly defined. In a video game, you are left to your own devices to work out what you can and cannot do, and this is where I think a lot of fun comes from.

You won't know it's an exploit till you exploit, so you can't blame people for looking for exploits because I think tha'ts what we're supposed to be doing in a game.

 

For me, every additionaly exploit I have to ignore makes me less immersed.

[oDDity]

And being able to escape from a guard by climbing a ladder doesn't?

I've already made claer that two attack animations are all that's needed to stop the AI being exploited on ladders.

As for the player putting a blockade at the ladder, I'd see that as a legitimate tactic rather than an exploit.

It'd be pretty hard to set up anway. If you've just climbed down a ladder, it's going to take you a while to get something over to block it with. You'd be as well just to carry on running and hide somehwere than hanging around making a noise. I don't think that alone is enough reason not to have it in the game.

THe player can just as easily blockade stairs - we going to stop the AI from climbing stairs as well?? What are we going to do then, give them jetpacks?

THe whole blockade thing is a complete red herring. IT would be a fun thing to do for a start, certainly a lot better than attacking the AI. Also, it's not going to be the course of action you would take 90% of the time. THere would probably be other guards after you as well if you were being chased, and if you block off the ladder you're then going to have to make a lot of noise removing the blockade if you want to use the ladder again. It just wouldn't happen much, any more then players blocking off stairs currently happens.

So, let's leave the blockade thing out of it.

THe only two questions are the AI needing an attack while on a ladder - easily sovled with two animaitons, and ladders not being set up proplerly - easily sioved by making prefab ladder templates.

[Springheel]

The best attack when chasing someone up a ladder is to grab the leg of the thief and pull him out and thus throw him to the ground. The best tactic when being chased up a ladder is to stomp on your pursuer's fingers.

 

We talked about the AI perhaps pulling the thief off the ladder, but that's not without its problems either. If the player is above the AI and falls off the ladder, he lands on the AI's bounding box. And it also doesn't help if the player is no longer on the ladder, but is simply standing in the way of the exit point. Still, it's a thought.

 

The whole point of the game is to exploit the AI's weaknesses to complete the mission.

 

Exploiting weaknesses that real life guards wouldn't have--like hopping on a table and knowing the AI can't do anything about it--does nothing but remind you that you are playing a game. Why do you think people want AI to climb ladders in the first place?

 

The only valid way I can think of for having guards climb ladders is that they only do it when the thief is not around. But that doesn't really solve any of the three reasons for adding ladder climbing in the first place.

 

THe player can just as easily blockade stairs - we going to stop the AI from climbing stairs as well?? What are we going to do then, give them jetpacks?

 

AI can push objects out of the way when they're walking, but that's not very feasible while climbing. And the thief can easily use themself as the block. Stand at the bottom of the ladder and the AI cannot get down. Parry his occasional swings and hack away at his legs until he is dead, since he can't parry you. This kind of problem doesn't exist anywhere else.

[bob_arctor]

In Tclassic AI go through crates, so no exploit there. AI go through the blockages...

 

Basically it comes down to level design, have 2 ways up, ladder and stairs, or make the AI use lifts and have them, and the AI will corner the player at the top.

 

Ladder climbing is hard and unecessary because if AI can bring out a ranged weapon WHY WALK INTO THE TRAP? An armed, dangerous intruder and you walk up the ladder to them????

Or find another way around and get the thief from behind?

 

I think though a cheat AI that climbs ladders out of view of player (Ie safe ladders, no risk of getting kicked/stabbed etc) would be fun though sometimes.

 

I suppose you could have simple TDS ladders for AI and harder ones like in chimneys for players, but then you would see a ladder and go "It's a non-AI one lets run up it!"

[Domarius]

And being able to escape from a guard by climbing a ladder doesn't?

Touché.

 

But we were going to rely on other ways to get around this - the guard waits at the bottom for you if there is no where else for you to go (instead of going to search mode like in the previous games "duuuh where did he go..")

And if there is another way up, send other AI to go around the long way to ambush you.

 

 

I'm going to think about what you said with the upper and lower attacks. The complexity doesn't end there you know. First, you need additional anims for mounting the ladder from the top, and another for mounting it at the bottom. Unmounting the ladder coudl be the same anim but in reverse.

Unlike the player, the AI needs to have body awareness. This means they ahve to get sucked on and off the ladder the way we did in T3 - this clumsy but nessecary approach will be exploitable.

 

If you beat the AI down to the ground and start shooting him, he could have an emergency "slide down the lader" ability to get to the bottom real quick and continue chasing you. But if you beat him to the top and start shooting him, he slides down, then he tries to chase you - back up the ladder again. Dumb behaviour. So maybe always run away if the player is at the top.

 

At the very least this is going to require modifying the AAS pathfinding system to include ladders in the process, and I don't think that will be a small task.

[bob_arctor]

It solves the 3 reasons as stupid people wouldn't follow blindly up a ladder while a dangerous intruder is there when they can

A) find a way round

get help

C) guard the exit point, the only one presuably except windows maybe(set new patrol there?)

D) Shoot from below

[FishFace]

So, let's leave the blockade thing out of it.

 

Crate blockades aren't so important - it's player blockades. The sweepy sword idea frankly sucks - as has been pointed out, the guard would just fall off. You still need contingencies, then for if the player is squatting at the bottom as the AI comes down - he's stuck there and what does he do? All this needs to be coded for. If he just goes back up, then that's unrealistically exploitable. No, no AI system is going to be perfect, but these are gaping holes in this bit of implementation. If it were a corridor, it would be like the guard would just freeze if he met the player.

 

What you say about self-restraint is utter twaddle. If the possibility is there, it's going to be niggling at you - "I could just crouch beneath him as he climbs down" for example. Even if you don't do it, the immersion is thrown to the pigs. Remember you're the one that advocates removing all weapons - if it were the same as having weapons and not using them, why do you want to get rid of them?

 

You still haven't said anything about nonstandard ladders - i.e. ones that require jumping in one form or another. These are quite common - common enough that preventing their use would be very annoying.

 

Sheesh, I'm glad you guys already decided on this one...

[Domarius]

Man this is a hot topic.

 

I think bob was replying to my first paragraph before I edited my post.

[bob_arctor]

My opinions still stand.

 

As long as good alternatives are considered like possibly unseen climbing, but more importantly

 

good pathfinding for other way round,

shooting the player,

staying guarding,

getting help,

 

those kind of more useful less specific things once the player does something different (jumping, mantling, climbing, these AI things apply to it all, the equivalent of Thief's "Damn You!!!" "You can't stay there forever!!!" "I'll get help!!" things).

 

Such a lot of work for one specific situation would be silly.

[Maximius]

Domarius the Wise:

In my opinion, the goal of any game where you have an opponent is to look for weaknesses in your opponent and exploit those weaknesses. In a real life sport game, the rules of what you can and cannot do are clearly defined. In a video game, you are left to your own devices to work out what you can and cannot do, and this is where I think a lot of fun comes from.

You won't know it's an exploit till you exploit, so you can't blame people for looking for exploits because I think tha'ts what we're supposed to be doing in a game.

 

For me, every additionaly exploit I have to ignore makes me less immersed.

<{POST_SNAPBACK}>

 

 

Bingo, Bammo, Right on the Money. When I played the Thief series, I used every damned thing around me to defeat the map andits AIs. To my thinking, if I can clear out a castle full of AIs or trap them or anything else to disable them, the FM author has failed, not me.

 

There was a T2 FM, a really great map of a prison, where you had to sneak in via a well in the basement and free someone. I ran around and around getting all theAI guards to chase me, then ran back intothe room with the well. Then I simply ran around and around the hole in the floor until the AIs fell in and drowned. I had a screenshot underwater with like 10 dead guards at the bottom of the well. All that was required to address this exploit was a lip around the well, but, well, OH WELL!

 

Another example was the FM Lord FishKills Curse. In that FM, a sorcerer has placed a curse on you , so that whenever your light gem gets so bright, a magical monster appears and tries to kill you. A great idea, but it backfired. Why? Because I would simply wait until a bunch of guard AIs were around, then run out into the light. The monster would appear, I would run to safety, and the monster would smoke the AIs for me, clearing a path whereever I went. After a minute the monster disappeared and I moved forward mostly unhindered.

 

Some exploits, like the Nudge and the Leaning BlakJaK I did not use because they were obviously flaws in the Dark engine. Nonetheless, they broke immersion for me just KNOWING I could use them but choosing not to.

[Ishtvan]

You still haven't said anything about nonstandard ladders - i.e. ones that require jumping in one form or another. These are quite common - common enough that preventing their use would be very annoying.

 

Actually the AI could probably handle climbing halfway up a ladder and jumping off at the top. The pathfinding was oddly enough based on Quake 3 bots who had to jump off of jump pads and do crazy acrobatic manuevers to pick up equipment in multiplayer maps, so the system is there to tell them to do that kind of thing.

[oDDity]

It's possible to set ladders up so they can't be blocked at the top. IF the ladder goes up through a trap door (which they usually do) just make sure the ladder goes well above the floor level in the upper room. There's no way the player can block it with his body, since there will be at least 3 diections to jump off at the top. THere could certainly be several templates set up that would aviod this probelm, FM authors just have to build their ladders following certain rules.

As for coming down the ladder, as domarius says the AI can have a slide animation for a quck decent. Whenever they hit anything on the way down, they detatch from the ladder.

[FishFace]

So you find a way to trigger the detachment?

 

If ladders can be implemented for AI without making them exploitable, and without making them sillily useless, I really am all for them. It's just the work:benefit:newbadness ratio that needs to be considered, and needs to be considered hard.