Whatever Happened to CryEngine?(lumiklovstad.wordpress.com)
lumiklovstad.wordpress.com
Whatever Happened to CryEngine?
https://lumiklovstad.wordpress.com/2019/05/19/whatever-happened-to-cryengine/
15 comments
This drew me in but didn’t answer questions it’s raising. Why is the engine so hard to work with? What about it makes it so specific to Crysis and why are they insufferable?
I only have some experience with Lumberyard right after Amazon purchased the CryEngine code base, so it was pretty much still 100% CryEngine (although an older version, IFIR CryEngine4). But still take my words with a grain of salt, it's all based on outdated information.
It basically comes down to that CryEngine isn't Unreal Engine or Unity. The editor looks different, the workflows are different, the way to implement things in code is different. If you have a team with good CryEngine experience (ideally people who had worked on Crytek games and the engine itself), then this isn't much of a problem. But if you put someone in front of CryEngine who had ever only worked with UE or Unity, they are lost and basically need to start from scratch (and at that point, you can just as well roll your own engine tailored to the game you're making).
There's been a lot of code and concepts in the engine that only make sense for first-person-shooters of the FarCry/Crysis flavour. For any other game this is dead weight at best, or requires the team to come up with hilarious workarounds at worst.
And the engine hasn't been "exposed" to many different teams and game projects as much as UE or Unity, so there are little travelled code paths that don't work as expected or just contain bugs, and the engine hasn't been brought into shape by external feedback.
Now all of this isn't really unusual when working with foreign massive code bases and tools, and game engine choice is often more about religion/ideology than pure facts (Stockholm syndrome etc). But in the end when you have a team that's familiar with UE4 or Unity, don't expect that this team loves to work with CryEngine (or Lumberyard).
It basically comes down to that CryEngine isn't Unreal Engine or Unity. The editor looks different, the workflows are different, the way to implement things in code is different. If you have a team with good CryEngine experience (ideally people who had worked on Crytek games and the engine itself), then this isn't much of a problem. But if you put someone in front of CryEngine who had ever only worked with UE or Unity, they are lost and basically need to start from scratch (and at that point, you can just as well roll your own engine tailored to the game you're making).
There's been a lot of code and concepts in the engine that only make sense for first-person-shooters of the FarCry/Crysis flavour. For any other game this is dead weight at best, or requires the team to come up with hilarious workarounds at worst.
And the engine hasn't been "exposed" to many different teams and game projects as much as UE or Unity, so there are little travelled code paths that don't work as expected or just contain bugs, and the engine hasn't been brought into shape by external feedback.
Now all of this isn't really unusual when working with foreign massive code bases and tools, and game engine choice is often more about religion/ideology than pure facts (Stockholm syndrome etc). But in the end when you have a team that's familiar with UE4 or Unity, don't expect that this team loves to work with CryEngine (or Lumberyard).
Ryse Son of Rome used CryEngine. Holds up to this day. https://www.youtube.com/watch?v=LgdFa7wT2ZE, https://www.youtube.com/watch?v=znYxqt4lUXc
The other big one is that the editor UX was completely bonkers which made learning it basically torture as there was essentially zero useful documentation. This is fine when you have or can build institutional knowledge which was often the case in the early licensed engine days but just made CryEngine woefully uncompetitive for modern audiences.
The Quora link mentioned in the article seems to have a bit more in depth information about why the engine is so difficult to work with: https://qr.ae/TWNufs
And then Amazon bought it and turned it into their lumberyard engine. The rumor is it is one of the reason their game crucible bombed.
Games bombing are almost never due to technical issues rather than people ones. For a game the size of Crucible it’d be my guess that the issues stem mostly from leadership and a team ending up thrashing trying to wrangle the game to fit changing visions and priorities. The end result of that is always going to be releasing something technically impressive but lacklustre.
Considering that valve are in at least some ways trucking along with the source engine I don't think you can really blame a game engine like that.
The engine usually gets blamed first though, no matter how good it is ;)
But yeah, Lumberyard... really wasn't good.
But yeah, Lumberyard... really wasn't good.
Why is the article talking about CryEngine as if it's not being used anymore? It is, plenty of games use it. The upcoming game The Climb 2 for VR is using it for example.
That's a Crytek title though. There's not many external teams licensing the engine (never really have been though). Arguably, focusing on creating great games instead of licensing the engine makes more sense for them anyway.
CryEngine was just that - an engine for Crysis games. Early Unreal engines also had the same issue of being an engine for FPS games and nothing more. Later on they transitioned to a more generic engine design and things like Unity3D caught on and no one really cared about CryEngine anymore (except for Star Citizen, but that's another story).
Unreal still has that problem. It's understandable but getting low enough into the engine to do weird shit is still very painful and basically undocumented.