Hacker News Reader

Comment by sidlls | Hacker News Reader

To underscore the sibling comment: at the point where a "% of revenue is millions of dollars" the company can easily afford to invest in an in-house solution if it's that important for their revenue goals. The tradeoff is hiring a sufficient number of skilled game engine programmers to make that work. It's not cheap.

View on HN

They don't even need to do that. If they are large enough where it would become millions there is a different license for them that they can negotiate with Epic.

Note: it doesn't have a royalty

This is what CD Projekt Red did?

Just guesswork here, but I think they have an agreement that is cheaper and also includes CDPR engine developers contributing code back to UE that makes it better to use for open-world games.

In the UE5 Keynote it seemed more like a technical partnership than just CDPR licensing UE5 for a price.

Also it is not just making a game engine but you also have to build some kind of a editor, script/animation tooling, plug-in support, etc.

Basically if you are big enough that you can make your own engine the engine also has to be usable by non programmers making the actual content for your game.

Tools can be nearly as important as the engine. I've read several postmortems where the lack of good tools made the asset creation and testing pipeline take way longer than necessary, leading to delays and crunch to compensate.

Tiger Engine that Bungie uses in Destiny supposedly took hours to rebuild a level and didn’t do incremental builds or cache so fixing that one bad light on the map was potentially days of wait.

Still better than earlier versions of Frostbite Engine from EA Dice that took week if no cache was present and ran on server farm that took raw assets together with metadata and spit out final game assets.

Reply on news.ycombinator.com