[CorpOverreach]

I wish more games would prioritize couch co-op modes over online play. Games that are focused on online only play basically have an expiration date from the day that they launch. Some may live longer, some may be dead on arrival.

But, make a good game that's playable by friends together at any time on a rainy day? If the game is good, it never dies.

[jayd16]

I think the cloud providers should figure out some kind of service for perpetual matchmaking/hosting of private servers. Devs are not always going to open source things but if you could fit your game server in some kind of package for Amazon to host then you can skirt that issue.

In theory, enthusiasts could pay to keep the lights on even after the developer went out of business.

[numpad0]

I would have totally agreed, but it looks like Steam offers free matchmaking backend and P2P proxies. Why isn't this widely known???

[Jach]

I would hope it actually is more known among developers. Even big titles like Tekken 7 used it. Insistence on crossplay with consoles causes a stumbling block, but still, that's not most games.

https://partner.steamgames.com/doc/sdk/api/example provides a good example, too. Steam has all sorts of extra functionality to justify their 30% cut on sales, they're not just taking it while doing nothing.

[Arelius]

It is well known. But also not all developers want to/ can afford the risk to tie all their networking services to Steam.

[Arelius]

Honestly, it's probably simpler than this, and the root of the problem often comes down to the Cloud Providers to begin with.

It's astounding the frequency I get an email from some cloud provider, or mobile app store that says something to the effect of:

"(Version X) of Dependency Y that we convinced you to use 5+ years ago is getting deprecated on August 1st, if you don't upgrade to Version X+5 you're service will go offline"

And we're stuck looking at the minimal amount of players running of that platform, and the hard choice of do we move precious human resources off of some in-progress game, that's already running late to learn a system that they never worked on, because the original people are long gone?

So, that's often why our network services, and mobile versions of our games are being taken offline while the single binary we shipped to one of the serious console vendors 10-20 years ago is still running, and now running on consoles 2 generations newer.

So, yeah, it'd be great if we could ship a package for Amazon to host perpetually, but first you could just get Amazon to care enough to ship a stable platform to build upon that wouldn't get depreciated.

[Aeolun]

Isn’t that just a matter of building your server with PHP or something similar that never deprecates functionality? Absolutely not node.js.

[Arelius]

Maybe...

But then, having also lived it, upgrading to newer versions of PHP and it's required modules is also not trivial.

And it's often not just your language, the cloud vendors have a lot of incentives to get you to use their hosted Postgres, or AuroraDB, or GameLift. Or even something as simple as we built all of our deployment scripts/images for our PHP system on Amazon Linux X, and now for reason Y, only Amazon Linux X+2 is supported.

[int0x29]

This isn't a technical problem. It's a legal and corporate political one. Copyright and patent issues are no more likely to go away with Amazon. Also Amazon's gaming division may make them a potential competitor.

I also doubt that such a could service would be immune to corperate restructuring by the like of Amazon. We need gaming companies to be more comfortable providing server binaries if we want anything that lasts.

[chatmasta]

The reason for the “de facto expiration date” is that eventually not enough people will want to play the game for matchmaking to be consistently available.

[Gamemaster1379]

Honestly Valve had it right with offering dedicated server packages. I respect any studio that does the same, like TripWire and Killing Floor.

I run my own private server for a live service game that shut down in just 1 year. We got lucky because they seemingly bundled the server code into the client. But the game was never meant to allow for that...

[valryon]

We did a couch coop game (CTHULOOT) and the number 1 refund reason we have is that it doesn’t have online. (It doesn’t because we lacked budget) So I’d say players now really expect online over local.

[janalsncm]

What you have described is survivorship bias. What matters is whether the refund request rate is higher than other games.

The people who enjoyed your game because it has couch coop (and therefore don’t request a refund) aren’t represented in that refund request stat.

[DontchaKnowit]

I mean, you can make both available.

[ChrisRR]

If you can afford to develop both. Presumably as an indie dev they chose where to invest their development time