Hacker News Reader

Comment by scrollaway | Hacker News Reader

I didn't expect this, but it makes a lot of sense in retrospect. CIG had already hired a bunch of the crytek devs and invested a ton into their engine. They're also already pretty tightly coupled with AWS.

I suspect this is the nail in CryEngine's coffin.

I massively disagree with the decision, though. I think CIG could have taken CryEngine under their wing and continued developing it, releasing StarEngine (their fork) as another competitor on a field which simply doesn't have enough high quality competition as it is. Maybe they thought about it; maybe they couldn't get the rights, or didn't have enough cash for it. Maybe Amazon made them a massive offer just for the name recognition, too.

View on HN

I don't see how fragmenting CryEngine benefits that much considering they're all on the same licensed CryEngine. Amazon paid $50m for the CryEngine and Amazon said Lumberyard is free to use with their AWS services.

I think it makes sense for CIG to use Lumberyard's engine with AWS integration, especially it saves them a lot of money not licensing their own CryEngine fork and they're using AWS anyway.

CIG probably has more crytek talent than crytek themselves. It wouldn't be fragmenting, at this point.

As far as I knew, the plan was to use the various Google Cloud services for the backend, but I wonder if CIG found the offerings from Google too paltry. AWS seems to have better geographical proximity for Australia and South America, along with Asia in general, which was a complaint/concern among backers from Asia and Australia and South America.

As far as the license for CryEngine, I believe that wasn't an ongoing cost per se. I know CIG has done a bunch of customization (64-bit floating point precision vs 32-bit floating precision the biggest) but they were originally talking about folding some of those changes back into the general CryTek offering. I wonder if they're instead going to offer that to Amazon for Lumberyard as a cost offset to their use of the services.

Star Citizen would be a pretty big feather in the cap for Lumberyard, as well. I'm optimistic about the game's development.

I suspect Lumberyard hasn't diverged from base cryengine much otherwise they wouldn't have been able to as quickly switch.

Amazon probably focused on code for autoscaling the backend server pieces and left the front end mostly alone.

That's my assumption as well. They have done massive amounts of custom work in CryEngine, so I'm guessing they are just using some features of Lumberyard.

A reddit comment [0]:

> StarEngine was based on CryEngine 3.7 (last big integration) with cherry picked changes from CryEngine 3.8 (mainly some Animation/Serialization fixes/improvements). The Lumberyard builds we used had the same CryEngine 3.7 base with the main changes on the Editor/Tools side. We heavily changed or have completely rewritten CPU/GPU side rendering, memory management, multithreading, entity/game object handling + added tons of engine features wich neither CryEngine nor Lumberyard have. Everything else is relatively easy to merge, thanks to the same code base + adding independent Lumberyard only features is straightforward too, especially with the help of Amazon developers. We are pretty good at handling the rest of the old CryEngine codebase, which CryEngine, Lumberyard and StarEngine have in common, ourselves ;)

[0] https://www.reddit.com/r/starcitizen/comments/5k06v9/letter_...

As someone who actually wants the game to come out, I'm glad they didn't.

Reply on news.ycombinator.com