So that Koster talk: old man yells at clouds? Engines (that do great rapid-prototyping) as blocking innovation and fixing cost/data seems off. https://www.raphkoster.com/2018/01/17/the-cost-of-games/
I feel like I was already hearing this type of sky-is-falling talk in 2003 when working on a Torque-powered game. The cost of data stuff feels weird to even include considering what constraints exist to shape how big a data set is once baked (imagine in No Man's Sky baked more assets, it could be a 150GB game if they wanted for the same dev cost).
@shivoa The engine part of that article strikes me as misguided. There's no per-engine trendline for costs, you can totally do procedural asset generation in a commercial engine (just have to write tooling for it, same as for an internal engine), there's no data there to support the claim.
Saying that the asset fidelity arms race is untenable would be true (and that's kind of what the cost/data chart shows), but that has very little to do with engine choice.
@shivoa that's just one thought though, could be correct or not. A question of "did N engines getting super popular stop us from investigating alternative, possibly better, approaches?" is a valid one. Perhaps I shouldn't say this as being from Unity itself, but I think the question *is* valid.
(of course a bunch of engine-haters did pick up exactly this point out of the whole thing and RT'd it loud & clear... but eh, internet's gonna internet)
@shivoa Yes. Also, data set is fairly small and could be very biased (lacks many mobile/non-AAA data points). That said, I'm happy Raph is doing this; I think we need more analysis of industry trends etc that are focused on development side, not just "this console/platform will win".