Automation/Scripting Interface (the way forward)

forman · 2021-06-03 10:12:51

I think a lot of the emergent problems players have been mentioning on the forums can be solved by varying degrees of automated behavior.

This could start with simple things like 'speed walk' to 'location' or 'forage' 'item' within 'coordinate loop'.

Some of this can already be done with custom modding which I will be looking into.

Potentially, this feature could be integrated in the game directly through crafting.

Relatedly, my main concern with modding is that the game is written in C++ and when making even simple changes, while dealing with manual memory, it is very easy to break things.

To that effect, I think that integrating some kind of api with a scripting language would make this easier.

I have not done this before but I am familiar with Python and I believe it would work.

All input is appreciated from a gameplay perspective, but I am particularly interested in what current modders familiar with the source code think.

Gremlynn · 2021-06-03 10:52:56

There are a handful of people who have worked on bots and ais for ohol in the past. I would ask around on discord and see if you can link up with them.

forman · 2021-06-03 12:09:14

Can someone give me a link or invite or whatever to the discord? Every link I've tried takes me to a page where I can't see anything.

Gremlynn · 2021-06-03 13:08:30

https://discord.com/invite/k4eU9CG

forman · 2021-06-03 13:57:55

expired I think

Gremlynn · 2021-06-03 14:40:55

Just used it and it worked fine. Maybe try using it within the discord app.

Morti · 2021-06-03 18:26:14

Rather than mod the experience of the game we share, I'd like to encourage anyone with that talent to instead make their own variant of the game, like Two Hours One Life and not use mods or any other clients besides Jason's when connecting to Jason's servers.

Mods take away the challenges that make the game difficult.
Zoom, search, threat and global communication mods all take away incentives the game is made to provide.
Those who have them won't cooperate with those who don't, on the same level.
Towns don't need to be organized if players have search functions, leaving new players to have to deal with the mess of items those with mods have no trouble with, and so vets have no incentive there to overcome that challenge.
Town layouts that are worked on by half or more people with zoom mods, are not built with the restraints in mind, those without have to deal with.

Adding automation to the world will take widen the gap between new players and those with the scripts to run them.

If you are gungho in doing this, just keep these things in mind. We want new players to feel needed, not in the way.
Mods, in the setting of this game, take away from it, they don't add to it.

forman · 2021-06-03 19:07:58

The way I see this working best would either be with some kind of 'drone baby' that a mother can optionally have or some kind of craftable robot. In either case, the bot performs one of several pre defined behavior patterns. Ideally, there would be some in game or mod interface that allows the player to define different behavior.

I don't ever expect there would ever be drones getting oil or doing trade -- but you never know.

But these features would enable greater in game projects. Families would not die as often (potentially not at all).

Players could actually venture out and create their own towns on their own with drones that other human players can interact with.

If Jason did want for there to be war-like gameplay (hence, the sword) this one of the only ways this would ever be viable.

And finally, whether using pre-built scripts or manually built through the game or modding -- maintaining a village of bots would not be easy.

Any type of population simulation requires a delicate balance of birth vs resource. OHOL, in particular, would be vastly complicated even if the player is there to manage the drones.

This would offer a whole new level of gameplay for vets who have run out of things to do and still requires intimate understanding of existing game mechanics to work.

I see this most likely being implemented in 2HOL, or possibly my own fork. In either case, I believe some form of automation is the direction the game needs to go to continue developing forward and also to sustain itself.

There are simply not enough players in general and also not enough good players to fill all the roles so that the game can run smoothly.

Drones are, in my opinion, the best (and probably the only) solution.

Gremlynn · 2021-06-03 19:30:56

As much as its a cool idea. It would likely break the game and the use of bots would be promptly banned. As morty suggests its probably a better idea to create a variant of the game with bot support like you suggest. Players could rule an entire town or city and go to war with one another. Two people but hundreds of bots. Though i do see a system like this getting very boring after a while. Also there would likely be a lot of server weight with that many bots running, so you would need to rent a monster of a server. And running it 24/7 wouldnt be cheap.

DestinyCall · 2021-06-03 22:43:34

Personally, I am not fond of using bots to solve our problems, even if it could.

A lot of the magic of living in an OHOL village comes from knowing that each villager is a real person. That everything in the village was made by human hands and as a direct result of other people's choices, for better or worse.

Someone planted this berry patch before I was even born. Someone crafted the hoe that I used to tend the carrot patch. This building was laboriously constructed, one piece at a time. Every piece of clothing or hot meat pie or hunk of iron was touched by human hands. And my contribution will become a part of that legacy of work and choice.

I don't want my neighbor or my own mother to be a bot. Or to give birth to bots. That's just ... eww. No, please don't.

Plus it would really ruining my immersion in the game if I have to check everyone I meet to confirm if they are a real player or just another robot. And it would make encountering a real player kind of unsettling, since they are a potential threat, instead of just a person-shaped tool.

Which is not to say that I am against automation in OHOL. I don't want to see it become Factorio, but I would welcome additional workbenches and tools of mass-production that would actually make late game towns feel more civilized. Upgraded versions of the adobe oven/kiln/furnace. More ways to make clothing and backpacks. New options for producing important materials, like rope, clay, stone, and coal. Additional ways to process raw resources into useful items.

The tortilla making table and loom are examples of what I want to see. But applied to other aspects of the tech tree, so we can have different cottage industries, like carpentry, shoe-making, masonry, etc. Brick laying might be fun.

Basically, better ways to make stuff, if you invest in making the necessary infrastructure to support more efficient production.

forman · 2021-06-03 22:44:14

It seems you are thinking of bots in the context of a game designed for bots and OHOL is not that.

Drones, in this case, are still bound by the same rules as players. If they are robots, they need fuel and if they are humans they need to eat.

This makes the cost of a drone extremely high and therefore there would be very few.

More likely than not people won't even be able to get them to be beneficial.

The notion of a town full of bots would require mastery of both game mechanics and scripting -- anyone at that level is surely not a griefer.

Gremlynn · 2021-06-03 23:08:05

You would be surprised forman

Bots would have the ability to scan the surrounding area for food, calculate the shortest path and the exact amount of time needed to eat to optimize food. A bot would also be able to interact with things on the server as fast as the server could register the actions. Obviously it would take a decent amount of scripting but simply calling them drones would be an understatement. They would be monsters no player could ever compete with. One would be unstoppable, now imagine a village full of them.

Basically This

forman · 2021-06-03 23:51:02

I'm pretty sure there are limits to how far anything can scan. For example, large portions of the map are suspended when no players are there.

To be able to eat efficiently you have to yum. Without yumming, a small number of bots would consume all of the wild food in an area very quickly.

That in itself could be problematic but the point is that to have the pips in order to do anything substantial you need a variety of food which requires a variety of cooking and resources.

This also requires an oven and tools which also requires a kiln steel, smithing etc

If you can script such a drone, then I dive hat to you, sir.

Gremlynn · 2021-06-04 00:03:37

Same to you.

Bots would be able to scan the entire render distance surrounding them for tools and resources. With the ability for super complex interactions to basically happen instantly food would never be an issue once a certain threshold was reached. Im not interested in programming an ai, ive got a lot on my plate atm already. That said I am aware of their potential.

Villanelle · 2021-06-11 06:53:53

If you bot any game completely it then plays itself. There is no game to play left.

Arcurus · 2021-06-11 20:13:04

forman wrote:

All input is appreciated from a gameplay perspective, but I am particularly interested in what current modders familiar with the source code think.

What i think about the source code, its worse then hell.... at least the server code... i did not look at the client code yet...

There are mods out there that allow to go in-between client and server, where you can then automate stuff without messing around with other code...

Only thing you would need to change in client is, that its accept external move commands for his own player. Vanilla does not allow it, so you can only trick it to believe its carried by another player...

forman · 2021-06-11 23:02:50

Gremlynn wrote:

Im not interested in programming an ai, ive got a lot on my plate atm already. That said I am aware of their potential.

and...

Villanelle wrote:

If you bot any game completely it then plays itself. There is no game to play left.

I work on procedurally generated/ai driven content and, believe me, its hard. It would break repeatedly and probably never be able to sustain itself without constant player corrections.

Arcurus wrote:

What I think about the source code, its worse then hell....

I was looking at the game source a bit and it seemed fine.

I don't see much point in self automation but it would be a good place to start as proof of concept.

My issue is that, if I was going to do this, I would want to do it properly with a solid scripting interface and this would take a lot of time/effort.

I don't see the point in investing without any guarantee that it would be implemented into the game.

I may talk with the 2HOL team at some point to see if they are interested.

Arcurus · 2021-06-12 11:13:17

forman wrote:

Can someone give me a link or invite or whatever to the discord? Every link I've tried takes me to a page where I can't see anything.

@forman
Im currently writing on a server based framework as server rewrite in haxe.

If you want to have more information you can contact me on discord: Arcurus#8211

Last edited by Arcurus (2021-06-12 11:30:13)

One Hour One Life Forums

#1 2021-06-03 10:12:51

Automation/Scripting Interface (the way forward)

#2 2021-06-03 10:52:56

Re: Automation/Scripting Interface (the way forward)

#3 2021-06-03 12:09:14

Re: Automation/Scripting Interface (the way forward)

#4 2021-06-03 13:08:30

Re: Automation/Scripting Interface (the way forward)

#5 2021-06-03 13:57:55

Re: Automation/Scripting Interface (the way forward)

#6 2021-06-03 14:40:55

Re: Automation/Scripting Interface (the way forward)

#7 2021-06-03 18:26:14

Re: Automation/Scripting Interface (the way forward)

#8 2021-06-03 19:07:58

Re: Automation/Scripting Interface (the way forward)

#9 2021-06-03 19:30:56

Re: Automation/Scripting Interface (the way forward)

#10 2021-06-03 22:43:34

Re: Automation/Scripting Interface (the way forward)

#11 2021-06-03 22:44:14

Re: Automation/Scripting Interface (the way forward)

#12 2021-06-03 23:08:05

Re: Automation/Scripting Interface (the way forward)

#13 2021-06-03 23:51:02

Re: Automation/Scripting Interface (the way forward)

#14 2021-06-04 00:03:37

Re: Automation/Scripting Interface (the way forward)

#15 2021-06-11 06:53:53

Re: Automation/Scripting Interface (the way forward)

#16 2021-06-11 20:13:04

Re: Automation/Scripting Interface (the way forward)

#17 2021-06-11 23:02:50

Re: Automation/Scripting Interface (the way forward)

#18 2021-06-12 11:13:17

Re: Automation/Scripting Interface (the way forward)

Board footer