Well, ok, you grasped at a few issues there that go off in different directions.

The issue with "Stop Killing Games" is that the legislation doesn't currently look like anything, it's a broad appeal and the solution for studios will depend on where it finally lands.

If it lands in the realm of "Games must be released FOSS after x years" then, aside from the fact that a lot of the times we don't own the copyrights to our own assets or certain code (they're on license for a single release) the second issue is how to release it.

First: the online backend for The Division or Destiny are just... not possible to run. The backend is fused to the products via a slurry of certificate pinning and object serialisation, with some things happening only on the server.

"Un-fusing" them is, basically impossible at this point; so the question is: can you build such a system without them being fused together in the first place?

The answer is: yes, but only by slowing down development. It would become much more about defining our boundaries and working on a "slim" version of the backend, or stubbing the backend completely. Obviously this is a lot of effort. The thing is we only barely managed to get a functional system, so adding an extra year for programming isn't going to be possible, we'll have to "cut" features that are hard to make.

"So, why don't you just release the server".

Well, that's a good question, we could remove the certificate pinning we have on the client, and the entitlement checks, stub out all the code that relies on third party APIs and give you a server binary.

But the server binary doesn't start unless you have 190GiB of RAM and 38 available CPUs.

So, we'd have to work on slimming that down, or building things in a totally different way: which means "seamless" darkzones and safehouses becomes impossible.

THEN you have the issue of releasing a binary that can be used to create cheats against the next version of the product, which we already had a major issue with.

So, most likely, we just make single player games.

Honestly, the industry is moving that way anyway because unless you've been doing it for a while making multiplayer games is really hard from a game design standpoint and there's an ongoing operational cost which people are a bit too price sensitive to support.

That's why Massive released The Division 1 & Division 2 but then pivoted to doing single-player games like Star Wars and Avatar which only retains the most basic multiplayer elements.

▲

singpolyma3 an hour ago | parent | next [-]

> But the server binary doesn't start unless you have 190GiB of RAM and 38 available CPUs.

This doesn't seem like much of an obstacle? Can buy or rent such without too much trouble.

▲

ashdnazg 2 hours ago | parent | prev | next [-]

> But the server binary doesn't start unless you have 190GiB of RAM and 38 available CPUs.

As far as I understand that situation is accepted by the initiative. The requirement is not that it works on any specific hardware or software stack, just that it can theoretically work.

> a binary that can be used to create cheats against the next version of the product

Anti-cheat solutions aren't required to be released, and if there are bugs in the server, they might even be found and patched by the community.

▲

dijit 2 hours ago | parent [-]

What you're saying is true for the californian legislation, but not the EU which is currently being drafted (in a different direction) - nor the direction of the authors article, and like I replied in a sibling response: it's not like people would be pleased to get our binaries.

Second: anti-cheat itself is a fucking joke. A crutch, a last ditch hail-mary because we ran out of time to batten down the hatches or things were changed so often from the start of the project to the end that we couldn't add safety into the protocol design properly.

Exposing how our systems think about how you move, how you shoot, when AI ticks, when loot ticks, behaviour trees and how phase transitions are computed: gives an attacker a hell of a lot of leverage.

To put this into broader easier to understand terms: ask yourself why it's so easy to cheat in Unreal Engine games vs Battlefield.

It's not the anti-cheat. It's the complexity of digging through the engine and knowing what the memory is doing and what the server is doing.

▲

strbean an hour ago | parent [-]

> but not the EU which is currently being drafted (in a different direction)

Where can we find information about the direction the EU is going on this? AFAICT there has just been one meeting on the topic?

	▲	dijit an hour ago \| parent [-]
		There's been... a lot more than just one meeting. https://citizens-initiative.europa.eu/news/14th-valid-initia... https://commission.europa.eu/european-citizens-initiative/me... https://www.europarl.europa.eu/committees/en/stop-destroying... https://www.youtube.com/watch?v=QXdmoeaYZ9Y https://www.eesc.europa.eu/en/news-media/news/eesc-debates-e...

▲

F3nd0 2 hours ago | parent | prev | next [-]

Wow, thank you for the detailed answer! I understand your point much better now.

I still think ‘kills any sort of multiplayer games’ (what the other dev said) is a gross exaggeration, since you list some ways this could be made to work, but it sounds like some things would cost significantly more resources and need to be done differently. But hey, maybe that’s not necessarily a bad thing. (Plus, there are multiplayer games which aren’t quite as resource-intensive on the server side.)

	▲	dijit an hour ago \| parent [-]
		I think what I'm trying to explain is that we barely make it work by the skin of our teeth, and adding more requirements means fewer features. The extra point I made was that it's actually kind of costly to run these systems, and I promise you publishers would love to push that cost onto the community with community run servers (think: CS1.6) but the reason they don't is because developing systems that way takes much longer and cannot be properly secured (mostly due to cheating but also from an entitlement standpoint). So, I think either multiplayer games will get much more basic, with simple gameservers. No more large multiplayer RPGs. Or, there will be fewer multiplayer games, because it's even more risk in an already risky business.

▲

tayo42 an hour ago | parent | prev | next [-]

We used to have player run servers for years. Is it some lost skill to write software that way?

	▲	dijit an hour ago \| parent [-]
		It's not a lost skill. Spinning up a binary and replicating actors across two computers that both have a connection string to a server is.. for all intents and purposes: easy. Where it falls down is when you start to have complex interactions with AI that's serverside, or you have a dynamic world that changes based on player behaviour, or you have cross platform requirements, integrations with companion apps and above all: matchmaking. If you're a looter-shooter, there's a whole host of complicated interactions too. A game like Apex Legends could probably distribute their server binary, but if you require online, as in, not just a single match, but an economy- a dynamic matchmaker and a dynamic world (meaning: when you kick a box it stays kicked) and a persistent account (you keep your loot): then that doesn't work well anymore. The interactions are just too complex to batten down reliably, they'll be exploited, there'll be no fun, or: it just won't be possible for certain features, regardless of safety. You can see how this looks by trying to use one of the many unofficial versions of Runescape.

▲

fragmede 2 hours ago | parent | prev [-]

> the server binary doesn't start unless you have 190GiB of RAM and 38 available CPUs.

> So, we'd have to work on slimming that down

...why? My reading of the law is that you need to make the binaries accessible, you don't have to provide the hardware to run it on.

▲

dijit 2 hours ago | parent | next [-]

Community backlash will be fierce if it's not actually runnable.

Ubisoft doesn't have the most stellar reputation for example (I don't work there anymore) so people look at things we do by accident as if they are intentionally malicious.

Also, the California law is one law, the EU is also looking at this and it's likely to look different - that's why "Stop Killing Games" doesn't really mean anything yet, even people within the movement have differing definitions.

	▲	ashdnazg 2 hours ago \| parent [-]
		The key is communication. If the company says the binary has a certain min. requirement, then the vast majority of people will accept that. Of course there'll be idiots, but I doubt you'll see a stronger backlash than to a company shutting down the servers without any solution, like they can do now.

▲

john_strinlai 2 hours ago | parent | prev [-]

>My reading of the law is that you need to make the binaries accessible, you don't have to provide the hardware to run it on.

if no one can run the binaries, despite them being accessible, then the regulation has failed and there will be a new movement to alter the regulation.

the spirit of the law is that i can reasonably spin up an instance of the server for me and my friends to play.

▲

ashdnazg 2 hours ago | parent | next [-]

Kind of depends on the definition of no one.

If the company puts an artificial proof of work demanding a rack of the latest data center GPUs, that should be illegal.

If the binary has the same hardware requirements that the company used when the service was up, I see it as totally fair.

	▲	john_strinlai an hour ago \| parent [-]
		true, but i think this would be exceptionally difficult (if not impossible) to enforce. ubisoft would surely be willing to spend an extra $500k on server hardware while developing a $25MM game, and subtlety bloat their server-side code so that they can say "this is the hardware we had to use to run it". there are a million ways to slow down code/increase hardware requirements that look plausible.

▲

strbean an hour ago | parent | prev | next [-]

If a game is popular enough for anyone to care, some turbonerd will get the server running on a massive cloud instance, and then people will be able to play the game.

Fans have reverse-engineered and stood up servers for tons of games with no access to the server binaries. The idea that they wouldn't figure it out when given much better resources (server binaries or source code) is crazy.

	▲	john_strinlai 15 minutes ago \| parent [-]
		>The idea that they wouldn't figure it out when given much better resources (server binaries or source code) is crazy. i wasnt implying they couldnt figure it out. i was implying that you would have to be an incredibly rich turbonerd to stand up a massive cloud instance for some of these games. which sort of defeats the entire goal of the regulation.

▲

stale2002 an hour ago | parent | prev [-]

> if no one can run the binaries, despite them being accessible, then the regulation has failed and there will be a new movement to alter the regulation.

This isn't the 2000s. People can rent a computer out of a data center. This isn't some hard problem here.

▲

john_strinlai an hour ago | parent [-]

>People can rent a computer out of a data center.

how much does 190GiB of RAM and 38 CPUs go for, hourly?

	▲	dijit 31 minutes ago \| parent [-]
		Cheapest I could find on AWS was $1.848/hr for the compute, no storage. $1,349.04/Month (m6g.12xlarge in us-east-1)