[strags]

> That way the community has the opportunity to run their own servers if they want to.

That might be fine for very small titles - where the "game server" is a relatively simple binary that can be run anywhere. Larger titles depend on a huge amount of infrastructure, for authentication, progression, matchmaking, etc... It's not feasible to open-source all of that, especially given that it may well still be in use for more recent titles.

[Ukv]

> It's not feasible to open-source all of that, especially given that it may well still be in use for more recent titles

If they're still running their authentication server (for example), then they wouldn't need to release that service.

Patching the game to no longer contact the authentication server would also be acceptable, for services that aren't a core part of the game. It's pretty likely the game already allows this for development/debugging.

If they've accepted money from people to buy the game, and don't want to keep the authentication service running, and don't want to patch the game to no longer require the authentication service, and don't want to refund people, and don't want to release the authentication service so others can run it - I think it's fair for a regulation to force one of those.

[knollimar]

So do games just have to have a perpetual endowment to fund any shared component costs? This seems like a logical conclusion. You wouldn't get scalability from reuse (e.g. reusing an auth library).

Or what's likely cheaper is budgeting for that patch in the game.

You may bemoan "oh they just don't want to release the auth service", but it functionally shuffles the cost math.

I'd personally rather the 5% cheaper games than trying to play a multiplayer only game 20 years later wtih 6 people on the server.

[Ukv]

> So do games just have to have a perpetual endowment to fund any shared component costs? This seems like a logical conclusion.

They don't need to keep services running perpetually. strags's objection seemed to be that it could be infeasible to release services like authentication that they're still running, to which I'm saying they don't really need to consider any of this until they stop running it.

> You may bemoan "oh they just don't want to release the auth service", but it functionally shuffles the cost math.

Releasing or patching it out is largely just fulfilling their side of the deal.

If I sell you a lawnmower that depends on some authentication server to start up, then shut down the server the next day (I got your money, why would I keep the cost?), and don't release the server code or a patch to work without it, then would you not say I've scammed you?

The resource cost of everyone I've sold to losing access to their lawnmowers would be far greater than what it'd cost me to release a patch, just that the former is not a cost borne by me if the law allows me to ignore it.

> I'd personally rather the 5% cheaper games than trying to play a multiplayer only game 20 years later wtih 6 people on the server.

Allowing a company to cut people off of their software (large cost) just to save having to push out a patch (small cost) will, on net, result on more expensive products - since on net you're wasting more resources.

Particularly when it comes to authentication checks, this doesn't just apply to multiplayer games. Imagine if this applied to other forms of media (already kind of happening with DRM), like if we couldn't read books from over 20 years ago.

[knollimar]

"If I sell you a lawnmower that depends on some authentication server to start up, then shut down the server the next day (I got your money, why would I keep the cost?), and don't release the server code or a patch to work without it, then would you not say I've scammed you?"

A lawnmower isn't a piece of software. It's not licensed. There's an expectation it should continue working. The lawnmower is a single player game.

I think it's understood that online games are a license (read multiplayer ones, not the Crew).

If I sell you an IoT lawnmower, and you get 20 years out of it, do I owe you a full refund if I shut down my server? imo, any refund should be prorated.

I get your externalities argument, but I think multiplayer games should be treated differently if there's not an easy solution.

>Allowing a company to cut people off of their software (large cost) just to save having to push out a patch (small cost) will, on net, result on more expensive products - since on net you're wasting more resources.

I don't think all the patches are small costs when you factor in licensing, etc. Also keep in mind if you use a library for networking and the API changes, do you have to then roll it on your own? I'm skeptical of the middleware that's made life easier

>Allowing a company to cut people off of their software (large cost) just to save having to push out a patch (small cost) will, on net, result on more expensive products - since on net you're wasting more resources.

This doesn't logically follow. Mandating you need to put out a patch creates a legal obligation that would sit on your books. Cutting the 12 people off your multiplayer game after 20 years isn't a large cost, and it's not going to make your next game more expensive. It was an externality that made consumers sad, not one making products more expensive.

>Particularly when it comes to authentication checks, this doesn't just apply to multiplayer games. Imagine if this applied to other forms of media (already kind of happening with DRM), like if we couldn't read books from over 20 years ago.

I think the DRM thing is a separate issue from the mandate for recoding games to be usable post server shutdown. I'd like it to be legislated separately as well. The books are analogous to single player games.

I think it's a slippery slope to turn that entitlement onto multiplayer games; if not that, then why not all software that you buy? Everyone should get a full refund when any software EoL's and companies go bankrupt whenever an online product stops being profitable.

You don't get a full refund in other sectors when they kill the consumable after a long while. Printer cartridges can stop being made and you don't get a refund for your printer. We didn't give HP the option "make your competitors ink work on your printer, give full refunds, support indefinitely, or open up your ink manufacturing line blueprints".

[Ukv]

> A lawnmower isn't a piece of software. It's not licensed. There's an expectation it should continue working.

I believe we should be able to have the same expectation of software, at least where not specifically sold as "X months of access".

> If I sell you an IoT lawnmower, and you get 20 years out of it, do I owe you a full refund if I shut down my server?

Ideally, to avoid unnecessary e-waste, you should patch out the requirement or release the server-side code so I can continue using my lawnmower. Buying it off me might also work, if you're offering approximately what I'd get out of having it continue to work, but I'm not sure if that scales well.

> I don't think all the patches are small costs when you factor in licensing

If bills like this pass, middleware providers would need to license under terms that allow distribution at end-of-life, or lose out on all customers selling software in California/EU/etc. Should also help clear obstacles for developers who already want to distribute their server/source code even before this law but are held back from doing so.

> Mandating you need to put out a patch creates a legal obligation that would sit on your books

There's no issue with creating the patch/releasing the server-side software early, just that I assume they'd want to maintain exclusivity to milk profit for as long as possible.

> It was an externality that made consumers sad, not one making products more expensive

If you expend resources to create some media/software/product, then brick that product while customers would've otherwise still extracted value in excess of the patch's resource cost (developer time, not licensing price), then you're on net wasting resources and thus making products in general more expensive.

Issue is that because the cost of patch is borne by the company, whereas they get to ignore the cost to the customers of bricking the products, the latter is often preferred even though it's typically the more expensive option by a significant margin. A bill like this should fix that.

> I think it's a slippery slope to turn that entitlement onto multiplayer games; if not that, then why not all software that you buy?

I believe it should, and that for a lot of software the case is even stronger.

> Everyone should get a full refund when any software EoL's and companies go bankrupt whenever an online product stops being profitable.

If you take someone's money to buy a CAD package, then no longer want to provide some service it relies on (usually just for authentication reasons), then you should release the server software or patch out the authentication check.

> We didn't give HP the option "make your competitors ink work on your printer, give full refunds, support indefinitely, or open up your ink manufacturing line blueprints"

I'd 100% support doing that!

[dontlaugh]

Plenty of games (especially MMOs) have lots of gameplay logic in the server. In many cases that is intertwined with the rest of the intrastructure, like databases, logging, deployment or even subscription services. Lots of games simply wouldn’t be functional without the publisher’s infrastructure.

Of course that is regrettable and could be changed, but it would require a significant change in incentives.

[strags]

Authentication is an interesting example - it sounds like might be the easiest component to remove. But without authentication, you don't have identity. And without identity you have no viable notion of accounts - and without accounts you don't have persistence, entitlements, progression, achievements, or any of the meta aspects that are deeply entwined with modern games. Not to mention how extensively identity ties into Matchmaking - another fairly complex backend service.

This legislation might be more persuasive if it were tied to a reasonable time limit, but I don't see anything of that nature in the text. An obligation to support or refund customers that lasts for a fair timespan (ie. preventing rugpulls) is far less onerous than an obligation to release your code to satisfy someone's nostalgia.

[Ukv]

For many games (and software, IoT devices, etc.), persistence/progression is tied to save files on your own device, and any authentication is more to the publisher's benefit of making sure you have an authorized copy - that's where it should be fine to just patch out the authentication check.

Even for games with centralized server-side progression, if you don't want to release the authentication service when you stop running it, then it could still be acceptable to patch out authentication by default and let those running community servers substitute in their own access system (like cracked/leaked unofficial servers already do).

> This legislation might be more persuasive if it were tied to a reasonable time limit, but I don't see anything of that nature in the text. An obligation to support or refund customers that lasts for a fair timespan (ie. preventing rugpulls) is far less onerous than an obligation to release your code to satisfy someone's nostalgia.

Should we be able to read books and watch movies that were released over, say, 20 years ago - or is that just satisfying someone's nostalgia? Maybe you don't think this matters for games, but with DRM it seems other media isn't far behind.

[account42]

Excuses. If there is a legal requirement you can watch all those concerns evaporate away.