Nothing quite as hilarious as a sea of people with zero experience using or writing "real" game engines arguing about the pros and cons of various specific features or engines.
On the upside, at least the people have finally picked up on the "the engine doesn't matter, the tools matter" meme I've been repeating ad nauseum. All the engine horsepower in the world doesn't matter if your content creators can't actually get any content running on the damn thing.
Few more points that I feel like bringing up in response to repeated silliness in this thread:
1) Engines don't make pretty graphics nearly so much as _art_ makes pretty graphics. There is no magic shader that just makes a game look good. Some of the best looking games (even in the last few years) have had little to no advanced graphics features in their engines, but rather made use of skilled FX artists and a solid art direction and cohesive artistic theme. Super shaders only matter if you're trying to make hyper-realistic looking games... which in turn still require millions of dollars in incredibly detailed and professionally made art which is generally out of reach of hobbyist game developers. Someone brought up Skyrim as an example of an engine that looks good: the tech in that thing is a steaming pile of crap. It looks good because artists put a lot of effort into textures, models, animations, and effects. The billowing snow on the mountains is art, not engine. The cultures and peoples of Skyrim are expressed via art, not engine. The magic spells and effects and monsters are all art and effects, not engine.
2) "Pretty" does sell games, like it or not. We might all like to pretend that only gameplay matters, but in the commercial world that is just not the case. There are a few games that have made many many millions off of fantastic gameplay and ugly as sin graphics (Minecraft, for instance) but there are many more games that have made hundreds of millions off of shiny graphics and boring repetitive gameplay (most sports games, fps games, eastern RPG games, etc.). The cash cows of the industry are all about graphics, and sell the umpteenth sequel by improving the graphics, not by changing or improving the game mechanics. Granted, these games also almost never use a premade generic engine... The idealistic "we just want good games" simply doesn't match market reality. Good games without good graphics can sell (and it depends a bit on what you call good; one man's "stylized" is another man's "utter crap"), but they'd sell better and easier with good graphics.
3) Graphics have an interesting curve to difficulty. You can go from basic to good relatively easily; you don't need a premade engine to pull it off at all, doing it from scratch is entirely feasible. Getting up to the super high end, however, is non-trivial. It very much can be easier to just start from scratch than to try to get an existing engine to do what you need. You can see this in companies like Turn 10 (you have to have something like a decade of experience and a PhD to even get considered for their graphics or physics dev teams; they are best in class, hands down) or even WB Games (who have rewritten the Unreal renderer -- supposedly the best there is, according to a lot of Epic Games fans -- because they wanted one that could do their specific graphics effects better than the stock one was easily capable of doing). Point is, if Torque were pure crap, improving it to industry norm would be easy. If it's already at the industry norm, the expectation that it can be taken up to an industry leader is misplaced.
4) Engines are almost always genre-specific, and the ones that aren't just suck for everything. Unreal is _really good_ at making Unreal Tournament and Gears of War. It's meh for everything else. Source is _really good_ at making Half Life and Left 4 Dead. It's meh for everything else. CryTek is _really good_ at making Crysis. etc. The generic engines I've seen tend to make things easy at the low end and are just a pain in the ****ing ass to work with at the high end. Unity, for instance, was pretty bad for anything but goofy little indie games until EA started licensing them back in 2010 or so and pushing them to actually fix all the things that were being a pain for their bigger AAA games. That said, even with the state of Unity now and the improvements coming in Unity 4, you still aren't ever going to be able to make the next Fallout or the next Minecraft in Unity, as there are always assumptions and limitations built into an engine that simply clash very hard with genres of games the engine is not designed for. That is one reason why you see a lot of FPS games come out of the major popular engines: they're all engines originally written for FPS games. This is also why I have to suppress the urge to backhand any developer that says "I don't know what game I want to make, but I'm planning my engine..." On that tanget, this article is always an excellent read for anyone considering to start a new game project: http://scientificninja.com/blog/write-games-not-engines
On the upside, at least the people have finally picked up on the "the engine doesn't matter, the tools matter" meme I've been repeating ad nauseum. All the engine horsepower in the world doesn't matter if your content creators can't actually get any content running on the damn thing.
Few more points that I feel like bringing up in response to repeated silliness in this thread:
1) Engines don't make pretty graphics nearly so much as _art_ makes pretty graphics. There is no magic shader that just makes a game look good. Some of the best looking games (even in the last few years) have had little to no advanced graphics features in their engines, but rather made use of skilled FX artists and a solid art direction and cohesive artistic theme. Super shaders only matter if you're trying to make hyper-realistic looking games... which in turn still require millions of dollars in incredibly detailed and professionally made art which is generally out of reach of hobbyist game developers. Someone brought up Skyrim as an example of an engine that looks good: the tech in that thing is a steaming pile of crap. It looks good because artists put a lot of effort into textures, models, animations, and effects. The billowing snow on the mountains is art, not engine. The cultures and peoples of Skyrim are expressed via art, not engine. The magic spells and effects and monsters are all art and effects, not engine.
2) "Pretty" does sell games, like it or not. We might all like to pretend that only gameplay matters, but in the commercial world that is just not the case. There are a few games that have made many many millions off of fantastic gameplay and ugly as sin graphics (Minecraft, for instance) but there are many more games that have made hundreds of millions off of shiny graphics and boring repetitive gameplay (most sports games, fps games, eastern RPG games, etc.). The cash cows of the industry are all about graphics, and sell the umpteenth sequel by improving the graphics, not by changing or improving the game mechanics. Granted, these games also almost never use a premade generic engine... The idealistic "we just want good games" simply doesn't match market reality. Good games without good graphics can sell (and it depends a bit on what you call good; one man's "stylized" is another man's "utter crap"), but they'd sell better and easier with good graphics.
3) Graphics have an interesting curve to difficulty. You can go from basic to good relatively easily; you don't need a premade engine to pull it off at all, doing it from scratch is entirely feasible. Getting up to the super high end, however, is non-trivial. It very much can be easier to just start from scratch than to try to get an existing engine to do what you need. You can see this in companies like Turn 10 (you have to have something like a decade of experience and a PhD to even get considered for their graphics or physics dev teams; they are best in class, hands down) or even WB Games (who have rewritten the Unreal renderer -- supposedly the best there is, according to a lot of Epic Games fans -- because they wanted one that could do their specific graphics effects better than the stock one was easily capable of doing). Point is, if Torque were pure crap, improving it to industry norm would be easy. If it's already at the industry norm, the expectation that it can be taken up to an industry leader is misplaced.
4) Engines are almost always genre-specific, and the ones that aren't just suck for everything. Unreal is _really good_ at making Unreal Tournament and Gears of War. It's meh for everything else. Source is _really good_ at making Half Life and Left 4 Dead. It's meh for everything else. CryTek is _really good_ at making Crysis. etc. The generic engines I've seen tend to make things easy at the low end and are just a pain in the ****ing ass to work with at the high end. Unity, for instance, was pretty bad for anything but goofy little indie games until EA started licensing them back in 2010 or so and pushing them to actually fix all the things that were being a pain for their bigger AAA games. That said, even with the state of Unity now and the improvements coming in Unity 4, you still aren't ever going to be able to make the next Fallout or the next Minecraft in Unity, as there are always assumptions and limitations built into an engine that simply clash very hard with genres of games the engine is not designed for. That is one reason why you see a lot of FPS games come out of the major popular engines: they're all engines originally written for FPS games. This is also why I have to suppress the urge to backhand any developer that says "I don't know what game I want to make, but I'm planning my engine..." On that tanget, this article is always an excellent read for anyone considering to start a new game project: http://scientificninja.com/blog/write-games-not-engines
Comment