Last week Meta announced that not only did it lose $4 billion in the Metaverse last quarter, but that it was also closing Echo VR (opens in new tab). The zero-gravity sports title was one of Oculus' early VR successes at the time, first on the Rift and later on the Quest headset, and Meta even acquired its developer, Ready at Dawn, in 2020.

The announcement of Echo VR's closure (it is still operational, but scheduled to close on August 1, 2023) has caused unrest among VR fans. In particular, it feels unusual for this to happen to a six-year-old game that, relatively speaking, has found an audience; CTO Andrew Bosworth has stated that "these resources can be used for other uses that we believe will be beneficial to the tens of millions of people who are currently participating in VR."

Interestingly, Bosworth also states that John Carmack, who has now left Meta, "would not have closed Echo VR." And how. Carmack sent a lengthy statement to UploadVR (opens in new tab) about the closure of Echo VR and does not seem impressed at all.

"I thought it was a mistake not to keep Oculus Rooms running and port it over to Quest," Carmack said, "and I thought it was a mistake to abandon all GearVR/Go content when my emulation layer was working at least to a large extent. I thought it was a mistake to abandon all of the GearVR/Go content. I believe in preserving everything."

"Even if you only have 10,000 active users, you should avoid destroying that user value whenever possible. Taking away something that is important to a user does more damage to your company than the benefit you would gain by providing something of equal value to that user or any other user. User value is my number one argument, but "focus" is also pretty high on the list.

Carmack also allowed the release of a root build of the Oculus Go (an unlocked OS that allows full access (opens in new tab)), which Bosworth had long pushed for, but "after seeing how much internal effort it took to make that happen, I almost felt bad," Carmack says.

"With a company the size of Meta, the constraints are different.

The co-founder of Id Software suggests several alternatives, such as having one developer maintain the games. As he says, cost-benefit analysis may not work for Meta, but "a lot of people are spending a lot worse."

Another alternative is to spin off the project: Meta would let go of the project, allow team members to leave, take over the rights for a fee of only $10,000, and keep it. This seems a bit far-fetched for a giant company that closes divisions like many of us drink coffee in the morning, and even Carmack admits that everything is "far from simple at Meta."

The other option is to leave the game unsupported but operational "rather than kill it outright," but as the game slowly rots, "that could lead to more of an online backlash than a clean kill." Open source is being considered, but Carmack admits that licensed commercial code would be a problem.

With apologies to Echo's VR fans, the reason this is more interesting than the only example here is that Carmack left Meta apparently dissatisfied with the way the company was run.

"It's true on one level that "it takes work to keep things alive," says Carmack, "but it's possible to build a system that runs untouched for years and still come up fine after a reboot. The default today may be a distributed mess of spaghetti, but it's a choice." 0]

And this genius coder advocates creating games that work "at some level" without central server support, encouraging LAN support for multiplayer games (because this allows people to write proxies), saving hosting costs Support user-run servers, both for the sake of it and the creative angle of the community. And, as befits a man who is currently devoting his life to rockets, Carmack takes off like a rocket.

"Be disciplined about the build process and what you put in your source tree. Think twice before adding dependencies that can't be redistributed." Don't do anything in your code that you wouldn't be comfortable with the whole world seeing.

"Because most game development is a panic rush to make sure things don't fall apart long enough to ship, it can be hard to devote time to basic software engineering. [The last phrase is a euphemism for "the game disappears completely and forever," and through Carmack's repeated exhortations and examples, we get the sense of someone who is very frustrated to see something built on a foundation that could ultimately prove self-defeating. We are entering an era where even many single player games require some sort of server ping, while others don't function at all when offline (which can feel like industry overkill outside of MMOs). And when it comes to games like Echo VR, which burns for a very short time and then disappears, it's hard to know if it was a strategic decision by Meta or just a rounding miscalculation: it just disappears quickly.