The California state assembly has passed the 'Protect Our Games Act'

Here’s a few, as someone who has worked in games for 12 years.

Most games have code and design decisions that hark back 25+ years. Every single unreal engine game for example is based code written in the mid 2000s and some parts of the engine really feel like it. Online components are developed the same way. If you made a multiplayer game 10 years ago and it was successful, your next game is going to be built on top of that. I’ve seen places that use stored procedures in Oracle DB for gameplay logic, others that rely on any number of SQL server specific tricks. Closed source dotnet frameworks, proprietary AWS services, if you can think of it there’s probably a game shipped on it. You’re also making the assumption that the server is a neatly coupled thing.

Am I responsible for providing a fallback to EOS, or Steam, or playfab in case their services are decommissioned?

What about the licenses for the code that affects other areas - we have a GPL’ed library here that we can use but now all of a sudden the vitality of the license means we have to replace it?

Who defines “ordinary use of the game”?. If the game has a multiplayer component, to some large number of users that can construe “ordinary use”. call of Duty is the best example of this (although COD is probably one of the games with the best track record here).

This is going to result in games moving more towards the Hollywood studio model - start up a company, launch a game and wind down the company for the next project. People who rely on that already unstable industry will be given even less stability due to this.

> I have a hard time imagining the server architecture would change much

That’s great - I’m sure if it’s that little work you’re willing to do it for all of those games companies.

> A dedicated company-owned server is just a beefier home computer with load balancers and matchmaking. Drop those two, slap a server list on the client, and you're golden

Game backends are just like Other backends. Some use event queues, microservices, third party APIs, licensed components. This adds a burden that no other software is expected to carry - it’s perfectly fine for Google to drop support for their devices but a 25 person company needs to go back and fix all their old games if they want to keep selling them?

> gives the game developers a perverse incentive to further embrace more exploitive revenue models such as free to play and subscription based services?

This is what I fail to see an explanation of anywhere in these comments. WHY would this law make a subscriber-based revenue model so much more enticing? WHY would this law make single-purchase games with multiplayer servers suddenly so non-viable from a business perspective?

The latent assumption I keep seeing is that the mere existence of a regulation in an area will drive people away from that model, but that's simply not how businesses operate. It's a cost/benefit analysis. So what is the cost?

There is lots of software not specific to games out there that comes with very murky licensing when it comes to redistribution. That’s fine because it’s predominantly used for backend, but now we’re talking about distributing binaries (or source code if you’re python). Here [0] is the license for a closed source python module - what do I do if I’m using this? SQL server is widely used and not redistributable, and has lots of proprietary features as a more common one

[0] https://github.com/Azure/MachineLearningNotebooks/blob/maste...

Voting with your money on its own rarely accomplishes much unless it is an overwhelming majority vote. There are also a huge number of things that simply cannot be managed properly, to the overall benefit of society, by demand-side market forces. There is a time and place for regulation, carefully considered and designed, with a change/revision process in place.

A big problem with lawmaking systems in the USA at all levels of government is that the change/revision process is virtually nonexistent. Laws are not adjusted as requirements change and understanding shifts. Regulation is hard in that environment, but the optimal amount of it is certainly much greater than "none at all".

It's simply not possible to maintain a functioning society without regulations on at least some things at least some of the time. Anti-regulation dogma is just propaganda by rich people who would become richer if their preferred bad behavior wasn't prohibited by regulation.

> I've seen some supporters of this style bill push for 'offline play' being a requirement.

That seems a bit silly to my eyes, self-hosting a server seems sufficient. But not included in this bill, so not an issue here

> Think if fortnite died tomorrow how many different servers it would take to host. Could an argument be made that an end user couldn't be expected to launch a dozen aws services? More dev time, more costs.

In this specific case, it's not so hard to imagine a single home computer handling the traffic of 100 connected users for a game of battle royale, the server compute for those kinds (baked-in world, low physics) games can be cheaper than running an instance of the game. Just some physics calculations, networking, and game state.

The main point would be if you start development from the premise that your server executable will be released to the users, the architecture/performance considerations are not that different at all.