Governance mods?

[ntnsndr]

Greetings. I'm new to this community, but have become quite obsessed with Minetest in a short period of time:) I'm exploring the idea of using it as a sandbox for implementing a research project about online governance, outlined in this paper on the idea of "Modular Politics." Basically, the idea is to create an API mod that could handle a diverse set of governance modules, such as for various kinds of voting, juries, and group roles.

I have two initial questions:

* What peer-governance mods are already out there? (So far I've just seen vote and its related mods.)
* Is anyone interested in collaborating on a project to expand the availability of peer-governance tools in Minetest (and potentially other platforms by extension)?

In case you're interested, here's a bit of a primer on my recent work about online governance.

[ywwv]

I want a bill of rights added to minetest. this could either be a mod or implemented into the game directly.

[twoelk]

at first I thought you were searching for tools along the lines of monitoring groups of pupils or such as in EDUtest or the classroom mod but I guess you are rather looking for stuff such as kick voting or sheriff voting.
As for a self-governing server the original Just Test server might have had enough drama and problems discussing to be an interesting source although it might be a long read.

When entering new multiplayer servers you will often have a wall of rules and tips presented near spawn. There are also mods such as Agree Rules or newplayer that will only allow players inside if they activly agree to rules shown in a message box. Sometimes you will need to answer questions to show you read the rules to get interact.

[ntnsndr]

Thank you so much for the replies!

@twoelk yes, you're right—more along the lines of the voting systems. And thanks for sharing the Agree Rules/newplayer examples. Those are definitely relevant. Are you aware of servers that use them?

@ywwv, can you say more about the bill of rights you'd like to see? It might be something I can include in my spec.

[MisterE]

If you want to simulate a governmental system on a server, then I suggest that you 'gamify' the idea. Create rewards, abilities, punishments, and credit systems within the scope of the game. You *shouldn't* mix this with server administration, except at a VERY low level. This is because Minetest servers, unlike the real world, exist because *someone* pays to keep the server up, maintains the mods, gets rid of errors, adds new content, and generally has the actual right and responsibility to maintain the server. From a little experience, I can say that being a server owner is no easy task or light responsibility. Therefore admin privs belong exclusively with the server owner, except those he considers safe to grant to others.

I would think that the highest level of admin privilege that could reasonably be given to non-admin players is kick. Perhaps only revoking shout and interact would be permissible for a non-admin government.

Be aware that players with the 'server' privilege, (by default only the system admin) can actually run code on the server's computer.


-----------------

so, the best way to make a government simulation experiment is to create abilities, rules that are supposed to be followed, but don't physically have to be, and clues for catching rule-breakers.

as an example, runs has made a pick-pocketing mod, that makes it possible to remove items from another player's inventory. This 'ability' was not available until then. He also implemented the chance of being caught... where a message is sent to the victim alerting that you are attempting to pickpocket them, and with the chance that your name is sent as well.

An example of what not to do is that the mod protection_redo (at least is supposed to) make it impossible for a protected area to have things removed from it or for it to be griefed.

On the other hand, the game node_core has no protection; furthermore, players inventories are accessible and possible to steal from, when they leave the game. However, it is really difficult to break blocks in nodecore, and for noobs, it is impossible to break stone (they don't have the tools), so a good house makes adequate protection against *most* thieves.

things to implement:
various forms of wealth, markets, economies
pickpocketing and theft
many levels of protection from theft, with appropriate cost of effort to install and use. No method of protection is completely invulnerable.
griefing and destruction abilities, with countermeasures (example: footprints mod allows crops to be trampled, but the mobs mod adds a mob fence that can't easily be jumped over)
offender detection systems (footprints (available), player detectors (available), player name recorders(easy to mod),etc)
transportation infrastructure, with appropriate cost in resources and time to set up and to use
ability modifiers (speed, jump, gravity, flight, noclip, weapons, etc) and complex ways to obtain and increase and decrease them

once you have done all that, you have a complex enough environment that you can set up governments to limit its use. You give the government enough tools that it can enforce its rules, even belatedly, and players *do* have the ability to build up enough resources to break those rules. Punishments and rewards should be set up.

One idea: have certain locations throughout the game that control of which gives the ability to rule a section of the world. Control of the location must be maintained to maintain associated local privileges/abilities. For example: if a clan named the policymakers gained control of a sweet spot, members of that clan all have a 50% damage reducer, 200% speed modifier, and 3 min of flight per in-game day, within a 3000-m range of that sweet spot. The sweet spot is claimed by the clan placing their flag within 5 m of the sweet spot, which has special particle effects generated at its location to mark it. If their flag happened to be removed from the sweet spot, then they would all lose the extra abilities they had. If another clan places their flag within 5 m of the sweet spot, then they too would receive extra abilities. To protect the sweet spot, policymakers build a thick wall of tough nodes all around the area, and place lava underneath the floor.
A nearby clan owns a sweet spot next door. They are called the X-rulers, and they want to take over the policymakers land. So they pool their resources, and get a legendary item that gives no-clip to someone for 30 seconds. One of the X-rulers takes the potion, then takes a X-ruler flag, and noclips through the policymaker's wall, and plants their flag. That gives enough of an advantage to their team that they can contend with the policymakers. But the policymakers have stored a legendary item that can revoke another clan's speak priveleges for 3 minutes. They use that and the other team can't coordinate. Etc Etc.

say that the priveleges that a sweet spot affords constitute a certian amount of 'power', that may be divided up among the members of a clan as they see fit. Certain members of that clan could receive 50% more power than everyone else, but that means there is less to go around. One member could receive all the sweetspots's power while everyone else in the clan receives none, or it could be split evenly. Also, rather than having 1 sweetspot per region that grants all the power, have many that each grant different types of abilities.

[ntnsndr]

Thank you so much, @MisterE. Such a generous response.

You really hit the nail on the head with emphasizing the centrality of who controls the server. If you're interested, I did a history of governance in online communities (ntnsndr.in/ImplicitFeudalism) that stresses, way back to the days of BBSes, how that model aligned with very centralized governance models. I could see, however, groups of people organizing and co-governing the servers themselves together; for instance, I'm part of such a group in the domain of social media—Social.coop, a Mastodon instance. If the server were co-owned, a kind of democracy of users would be a natural extension.

That said, the Modular Politics framework I'm working with (metagov.org/modpol) could just as well enable the server admin to have as much central power as they wish, while also allowing for users to adopt and even develop diverse structures underneath that.

The connection with economics is interesting and important, and I think it could play a role. Though on the whole I've actually found that overly tying governance to economics (as in blockchain-based DAOs) can be as much as a constraint as a benefit.

Your ideas about attaching self-governing groups to territories and privileges is absolutely something I'm exploring how best to implement. Thank you so much for fleshing out these ideas—they're super helpful.

[MisterE]

Thanks for your response to my ideas. I guess the main point of all that was that certain privs individual server owners will not want to and should not give up (for their protection even). However, If you build a fantasy world under that, that gives many more privs and abilities, then there should never arise a need for the community to govern the server itself... the community instead has its hands full with the fictional society. In that framework, as long as a user does not commit sins against the server (hacking, for example), they can't be banned. However, the community, in whatever form of government it has, can punish that user.
------------------------------------------------
more specific ideas:
(sorry, I have an active imagination :D)
-- One form of punishment could be banning to the outer lands away from central activity.

--using mobkit, you can make npcs with targeted behaviors. For example, A sufficiently rich government can hire 'policemen', The more policemen they hire, the greater a certain policeman spawn chance grows. Policemen have that specific chance of spawning near players that are on a wanted list, and they will attempt to apprehend (catch) the player on that wanted list. when apprehended, certain functions run that punish the player, according to what the governemt has assigned. Another type of npc could be mauraders, which have a greater chance of spawning in less-well-governed areas.

-- such a server really should disable global chat. Instead, have various communication devices that re-enable chat with various limitations. This is quite possible with chat callbacks. Alternatively, chat could be via formspec and HUD

-- in-game governments should not be all-powerful by any means. They should also be restricted in the geographical scope of their domain. There should be areas of anarchy and of order, and benefits and negatives to living in each type of area. For example, there could be a significantly lower chance of being harassed by other players or by "maurader npcs" within a settled well-governed area, but low access to ores and resources.

[ntnsndr]

Thanks so much again for this—the mobkit idea makes a lot of sense. And yes, the enforcement mechanisms are key here, and I have a lot to learn about how best to implement it.

On chat, one thing I've already been implementing is chat within a given "org," which could be either one's group or even just, say, that group's executive committee. But thanks for the tip around using formspec or HUD; I haven't gotten to the stage of working with the various interfaces yet!

[rubenwardy]

Hi, this explains your questions about vote in PM! Yeah, vote is designed more for automatic game voting with online players rather than actual democracies.

I was making a game for Minetest called Capitalism Game. I was intending to implement a democratic voting system here, for both companies and the government. People would submit bills which the shareholders/members of congress would vote on.

I didn't get that far into this. I implemented a lot of the executive branch - you could set CEOs/presidents and they could hire staff. I stopped the project because it takes a lot of time and it may not actually be that rewarding (ie: it may not actually be a good game)

[ntnsndr]

@rubenwardy, wow, that's really neat! I think Capitalism would be fun! Probably more fun than real capitalism:)

The approach I'm taking is less about designing a specific game than about creating some tools that would make it easier for people to build diverse games using diverse governance structures within them—companies, governments, pirate pacts, marriages, whatever. The dream is even to create an API robust enough that, if it's implemented in parallel on other Lua-extendable games, the same modules could be portable across games.

Another approach would be to adopt a tool like my colleague's PolicyKit.org, which uses an API with an external tool to enable governance mechanisms on platforms like Slack, Reddit, etc. Would be interesting to see if it could be bridged to a game like Minetest.

Where have you been seeing things like voting working well in Minetest in the wild? Do you see promise in developing this kind of stuff?

[ywwv]

hi ntnstr . sorry but the "at" symbol does not notifiy people which is hwy my response was delayed

I think it should be like the anabaptists . I don't beleive in an automatic social contract. You should have to join it like the amish. also enforcement should be physical. ethereal enforcement is too restrictive. the mod should be a framework to make governance possible- not a governor itself. there should be a system to try people and to bring people to trial- not just a legislature

[ntnsndr]

Thanks, ywwv. I didn't assume the @ worked to notify—I just used it out of habit:) Is there a way to actually push notification, other than for people to subscribe to the thread?

I agree on the automatic social contract. The way I'm currently building actually assumes simple consensus as the default mode for groups. They can quickly change that, but that's the norm, rather than founder autocracy.

And yes, as you'll see with the Modular Politics framework I'm using, the idea is a framework, not the governor. And there could certainly be various structures including trials, etc.

Can you say more about what you mean by "enforcement should be physical. ethereal enforcement is too restrictive"? I'm not sure what that means.

I'd love help from a more experienced mod dev, if anyone's interested!

[ywwv]

oh , I get it . "ntnsndr" is short for "NaThaN SchNiDeR" -> NTNSNDR... so you must have helped write that paper. pretty cool. is this a follow up on it? you don't have to conform this. I get it if you're trying to keep a low profile. it must be hard doing that if you have a wikipedia page.

there should have to be someone to physically enforce laws. like a sheriff. so instead of having a broken law automatically detected, an agent has to discover and enforce them. there should be items that have the power of the courts. then you can try people. by framework I meant framework for individual actions- not a framework for the government sysstem (I know it can do that already) . like a manager ... minimal physical action on the world. you might need do that to help with trials.

[rubenwardy]

The approach I'm taking is less about designing a specific game than about creating some tools that would make it easier for people to build diverse games using diverse governance structures within them—companies, governments, pirate pacts, marriages, whatever. The dream is even to create an API robust enough that, if it's implemented in parallel on other Lua-extendable games, the same modules could be portable across games.

Well, a game is a collection of mods - I had to develop lots of tools to allow companies and governments. However, the mods don't aim to necessarily be compatible with all game types.

Another approach would be to adopt a tool like my colleague's PolicyKit.org, which uses an API with an external tool to enable governance mechanisms on platforms like Slack, Reddit, etc. Would be interesting to see if it could be bridged to a game like Minetest.

My mods had a permissions system but being able to make a flexible system to manage this stuff isn't the easiest thing. Some things like voting for a new CEO/president are fairly easy, as there's a fairly simple outcomes. Things like stopping corruption and bad behaviour is a bit harder.

I suppose you could allow Lua scripts to be embedded in bills, which would act a lot like policykit. Real life uses a legal system to resolve disputes, this has the benefit of being flexible enough to handle edge cases. Code isn't usually flexible to human problems.

So perhaps a combination of automated permissions/policy checkers and some "legal" system would be a good approach.

Where have you been seeing things like voting working well in Minetest in the wild? Do you see promise in developing this kind of stuff?

I've seen servers that elect moderators and vote on proposed changes, this is typically done rather informally rather than with mods and tools. I think it's an interesting idea to explore, but I'm not sure how much it makes for good gameplay

[ywwv]

I think it's an interesting idea to explore, but I'm not sure how much it makes for good gameplay

Multiplayer sandbox gameplay can be thought of as a medium for interpersonal conflict- which is the real "game"- higher orders of societal organization are condusive to this. The content of the game itself is only valuable in it's ability to deliver the real product. Materialism is an abstraction of it's social significance- not its own end.

[ntnsndr]

oh , I get it . "ntnsndr" is short for "NaThaN SchNiDeR" -> NTNSNDR... so you must have helped write that paper. pretty cool. is this a follow up on it? you don't have to conform this. I get it if you're trying to keep a low profile. it must be hard doing that if you have a wikipedia page.

Oh, I'm sorry—I didn't mean to disguise myself at all. I use that handle everywhere. Not trying to keep a low profile. Yes, I'm an author of that paper.

[ntnsndr]

Getting the hang of how this forum works:)

Well, a game is a collection of mods - I had to develop lots of tools to allow companies and governments. However, the mods don't aim to necessarily be compatible with all game types.

Yes, in this case, though, I think it would be a palette of mods which people could mix and match into very different kinds of games.

My mods had a permissions system but being able to make a flexible system to manage this stuff isn't the easiest thing. Some things like voting for a new CEO/president are fairly easy, as there's a fairly simple outcomes. Things like stopping corruption and bad behaviour is a bit harder.

I suppose you could allow Lua scripts to be embedded in bills, which would act a lot like policykit. Real life uses a legal system to resolve disputes, this has the benefit of being flexible enough to handle edge cases. Code isn't usually flexible to human problems.

So perhaps a combination of automated permissions/policy checkers and some "legal" system would be a good approach.

[...]

I've seen servers that elect moderators and vote on proposed changes, this is typically done rather informally rather than with mods and tools. I think it's an interesting idea to explore, but I'm not sure how much it makes for good gameplay

Certainly some things are always going to make the most sense outside of mods/tools. But I'm also interested in what could happen if we could get our mods and tools up to snuff with the functionality we expect in offline companies, clubs, associations, local governments, etc.! Can't guarantee the gameplay will always be good, but I'm really interested in exploring where it might be.

Thanks all!

[sorcerykid]

On the JT2 server, which is predominantly anarchy survival, I've been developing a Voting Booths feature, which will allow players to submit measures for vote on issues and concerns of relevance to the community, such as auctioning abandoned properties or moving a controversial structure. The voting process is designed to ensure fairness as well as to mitigate abuse by requiring approval amongst a sufficient number of senior players and staff within a limited timeframe (both dependent on the issue at stake).

The inherent limitation of any in-game system of government, however, is that nearly all participants are volunteers, and thus they have no obligation nor expectation to be present indefinitely, which dramatically reduces their capacity to govern in a meaningful way. This means that some degree of automation (whether that be through use of AI, algorithms, heuristics, etc.) will be necessary to maintain social order and cohesion. Likewise, it puts greater pressure on the server administrator(s) to double as public officials, which can detract from their other responsibilities.

[ntnsndr]

On the JT2 server, which is predominantly anarchy survival, I've been developing a Voting Booths feature, which will allow players to submit measures for vote on issues and concerns of relevance to the community, such as auctioning abandoned properties or moving a controversial structure. The voting process is designed to ensure fairness as well as to mitigate abuse by requiring approval amongst a sufficient number of senior players and staff within a limited timeframe (both dependent on the issue at stake).

Thanks for letting me know about that! I'd love to try it out. Is the idea that you're holding the voting data in a particular object in the game?

The inherent limitation of any in-game system of government, however, is that nearly all participants are volunteers, and thus they have no obligation nor expectation to be present indefinitely, which dramatically reduces their capacity to govern in a meaningful way. This means that some degree of automation (whether that be through use of AI, algorithms, heuristics, etc.) will be necessary to maintain social order and cohesion. Likewise, it puts greater pressure on the server administrator(s) to double as public officials, which can detract from their other responsibilities.

Absolutely. The idea I'm trying to develop is to have forms of governance flexible enough that they can be made appropriate for the limitations and circumstances of groups. It really should be an opportunity, not a burden.

[MisterE]

one tool that might make sense in this context is a server lobby. This is where new players come when they join, and can be accessed at any time by entering /lobby

When players are in the lobby, they cannot be killed, and their regular gameplay characters are treated as if they were logged out in terms of survival gameplay elements. (e.g they do not lose hunger, health, etc.) You also cannot enter the lobby for 2 minutes after someone hurts. When you leave the lobby, you are put back into the world exactly where and how you were. Player state is stored and retrieved. You cannot bring anything into the lobby, and you cannot take anything with you out of the lobby. Anything that you have is stored and returned just like player state.

They lobby would be a safe space for players who want to hang out and use the server just as a social platform. It would also be a good linkage to other, non-gameplay elements of the server (example: server-wide meetings, personal dimensions, and to minigames, if the server has any. It would be a place for posting announcements, displaying a server map, having tutorials and museums of the various gameplay elements.

Functionally, the lobby would actually be a y-layer of the world, surrounded by unbreakable nodes. Probably It would be the top 100 nodes of the world, or the bottom 100. however, the actual min and max y should be settable per-server.

[ntnsndr]

Thanks for the ongoing discussion! Just posted a repo—very much a work in progress!
https://gitlab.com/medlabboulder/modpol/