Creepy Mannequin Update [Weeping Angel mechanics]

[zanity]

FO4 inherits from Skyrim, and so look at the range of Skyrim mods to see what is really possible and what is not.

 

The fine granularity control of actors in the game that you desire is NOT baked into the engine, so I don't think it can be done (as described). The helpful person above who gave you a potential script does what all modders do- map their desire onto how the engine actually does things- which usually gives you the best you can achieve.

 

The best way to mod a game like FO4 is the other way round to how you are currently thinking. Learn how the game engine does things, then be inspired to use that functionality in new ways. Seeking to impose 'fresh' 'code' ideas on this engine - things that could be trivially coded from scratch on a TRUE programmable engine- will usually lead to frustration and disappointment.

 

Look at 'true carriages' in Skyrim- a truly brilliant mod that built on a removed intention by Beth to have functional carriages for transport at one point in the game design cycle. The original Beth coder had a go of adding back this functionality with little success. A brilliant modder took on the task, and actually made something worth using. But VERY, VERY simple problems remained (like how the carriage interacts with people on the road) because of how much the actor AI is baked into the engine, and cannot be changed.

 

Most external ideas will be incapable of being 'perfected' as an ambitious FO4 mod, so perfection is not worth pursuing. Beth games need pragmatic modding where you roll with the punches- and lower your expectations of what the engine can do.

 

PS that's not to say you couldn't do a highly scripted localised version of your idea- like that mentioned in some other games with creepy moving dummies. It's just that turning your dummies into a general 'faction', and having them behave as those episodes that Doctor Who ripped off from better prior sources, is unlikely to fit how FO4 does the AI.

Edited May 7, 2016 by zanity

[zanity]

My first response was accurate but not directly helpful- perhaps I may have some specific useful ideas.

 

1) don't attempt to over-ride general mannequins in the environment- that is not a good idea for a whole host of reasons.

 

2) so firstly you need to place YOUR creepy mannequins for the player to encounter- and the advantage of this is you can choose times and places like the games that used the same idea to scare the player.

 

3) beginner programs make the mistake of trying to load all their desired functionality on a single engine mechanism/library. To them, that seems to be the 'skill' of coding. However, the different parts of your plan may best match different parts of the game engine. So think about having simultaneous systems active that you switch between to get all your needed functions.

 

So the mannequin is originally a static object. When the player is close enough and turns his/her back, the static object is removed, and an AI actor put in its place, with the required attack logic. If the player turns to see the mannequin before it is close enough to achieve 'attack', I think you want it to 'freeze' its pose. But the player hasn't seen its prior movement, so the frozen pose can be another static object put in place from your collection of such objects.

 

Making an AI actor 'vanish' before the player sees the 'trick' would be a matter of experimentation, to see what the engine does best. Perhaps instantly give the actor a high or low location on the map, or an insane vertical speed mixed with an 'intangible' physics flag (assuming the engine allows this).

 

After all, the next time the mannequin 'moves' it can be with a completely NEW created AI actor if that is more useful.

 

If the engine allows viewing angle of player vision relative to AI actor, you could 'freeze' the mannequin in an angle region BEFORE the player actually sees it, giving the engine some frames to respond to your new iinstructions.

 

However, this whole idea of 'freeze tag' doesn't really work once you have more than a couple of Mannequins, since befoe long they would be all around the player. That's why games that used this idea before used really fixed and limited scripts for the location of the mannequins.

 

You see- it will be FAR better to have most mannequins in the world ORDINARY so the player never know which to fear- and will let their guard down. If you want a bunch of them- best to accept that once they have really engaged the player in combat, the 'freeze' idea is flagged OFF- and they become a bunch of conventional AI attackers. Surviving attackers could have their static objects reset once out of combat range with the player.

 

Good luck with your mod.

[reelo2228]

@zanity

Oof, dayum! :D

Thanks for that thorough explanation, those were things i've never even thought about!

Given that i'm new at modding in general, i was fully prepared for setbacks and probable impossibilities.

While's fiddling around with the scripts, i had thought that might've been the case, but thankfully you just provided me with alternatives to ponder on!

(What is this better prior source you are referring to? :D)

Having read your analogy, i should rethink the way to approach this idea again, it kinda DOES feel like the game was fighting itself during the test, the AI is telling it to behave in one way and the mod is forcing a restraint (instead of immediately stopping, the enemy sort of went into a slow motion and froze eventually after 2 meters of sprint, was quite cool actually, like they were drenched in instant dry cement xD). But i'm hopeful that there's a workaround somehow.

I'm gonna crawl back to my cave now and try them out!

[Flyingdebris]

Dunno how well this might help, but feral ghouls do have the "laying around, feigning death" functionality where they basically lay around until triggered into waking up.

 

Maybe that can be used as a starting point?

 

However, should be noted that in a game where players can routinely carry around miniguns, auto shotties, and large quantities of ammo, having something that freezes when you look at it might be a tad bit too easy to deal with if you can just keep your eyes on the threat and calmly line up your shot. It may make better gameplay sense to use the weeping angel mechanic on something you can't really kill, like an obelisk or something (maybe down in dunwich borers). where your only option is careful avoidance. Mannequins might make more sense as something that simply may or may not come to life, and just straight up come at you conventionally. (you could justify it as disguised gen 1 synths)

 

I am reminded of a part of the old system shock 2 game, where up to a point in the game you've been dealing with reprogrammed c3p0 type robots that seek you out to explode on contact, polite all the while. And then you come across the warehouse where they are stored, where the walls are lined with the containers they are shipped in. Periodically they randomly start bursting out of some of their boxes. One of the most nerve racking parts of the game IMO.

[Reneer]

So i've tried out the scripts and here are the feedbacks \:D/

 

1. I've attached the AdderQuest script to a Quest in the kit, and ticked the box "start game enabled" and "run once", but in the test room i created, the object mannequins were not replaced. I read through the the scripts line by line, and although i don't understand every variable there, it all seems right and makes sense to me xD

Send / PM me the ESP file you created and I'll have a look at it. My guess is that setting "run once" on the quest is the problem.

 

2. It seems as if the game is too slow to respond to the first Actor script, i placed it on the Mannequin actors. Initially, they stand still, and upon detection, they run to me, then slow down half way and eventually stops moving.

Then, when i turn around, for a good 3 minutes or so, they start moving again, and this time, the script don't activate anymore, the mannequins begins to attack like a normal FeralGhoul AI (i used their attack data and AI as template).

Ok. I'll look into this too. If you gave them the FeralGhoul AI, they will attack like feral ghouls. I'm not sure what you mean by this being a problem. How are they supposed / meant to attack?

 

Assuming the problem is the detection code (which I doubt, but it could be the case), there are other ways to imitate similar behavior, such as using angle headings and such to fake detection / LOS.

Edited May 8, 2016 by Reneer

[reelo2228]

Dunno how well this might help, but feral ghouls do have the "laying around, feigning death" functionality where they basically lay around until triggered into waking up.

 

Maybe that can be used as a starting point?

That is one way of doing it, i also remember the Synths pretending to be a mannequin in the game at one point, i shall look into it and find out how it ticks!

The system shock 2 scenario was freaky for me too xD it was awesome!

 

 

 

 

2. It seems as if the game is too slow to respond to the first Actor script, i placed it on the Mannequin actors. Initially, they stand still, and upon detection, they run to me, then slow down half way and eventually stops moving.

Then, when i turn around, for a good 3 minutes or so, they start moving again, and this time, the script don't activate anymore, the mannequins begins to attack like a normal FeralGhoul AI (i used their attack data and AI as template).

Ok. I'll look into this too. If you gave them the FeralGhoul AI, they will attack like feral ghouls. I'm not sure what you mean by this being a problem. How are they supposed / meant to attack?

 

Assuming the problem is the detection code (which I doubt, but it could be the case), there are other ways to imitate similar behavior, such as using angle headings and such to fake detection / LOS.

 

Well, initially i wanted the actors to attack player only when he is not looking, when the actors are out of view, that's when they attack, and freeze in motion when they are within Line of Sight.

 

 

Another approach i'm thinking now, may suppose change the direction of what i initially imagined the mod to be. But...

So for gameplay sake, I could place a certain number of creepy mannequins in specific locations and have them engage the player head strong and viciously. Of course, somehow they should ignore when the player first enters a room, and only engage after a short time frame.

 

 

Edited May 8, 2016 by reelo2228

[Skyler6669]

Is this mod still in the works? Because it's probably one of the top mods ever made for a game. Very amazing idea.

[Demeggy]

Good work - this is how I got the Weeping Angels working on Fallout Who Vegas!

 

If you've got no objections, I'm going to have to hack at some of that example code, as we're steadily reworking Fallout Who for Fallout 4 at the moment, and this is just the ticket.

[jkruse05]

It would be cool if, at the very least we could get some mannequins that turn to face you when you're not looking.

[Demeggy]

In theory, very easily achievable for standard objects. Little trickier in principle this mechanism... been working on a rebuild of the example script provided; lots to consider - AI Packages being executed, LOS angles arent always sh*t hot either, so its quite ambiguous determining what actually LOS is etc.

 

Will keep you posted with findings.