3 ms·

Not to be snide but how many games have you shipped like that? Its really a massive undertaking to build a game engine. Not to mention a distraction as well.

by jayd16 3y ago

JohnFen 3y ago
You don't need to build a game engine, though. Just build a game. Game engines exist basically to be able to reuse code across the development of many games. If you don't worry about writing an engine, and just write games without one, then as a side-effect you'll also be developing the core of an engine that's more specifically optimized for the sort of work that you actually do. Eventually, it may become advantageous to make that into a formal engine, or it may not. Either way, you're still creating games.
- SXX 3y ago
  What you describing is only possible for hobby game developers with extremely small scale of a game or development time of 4-5 years. Majority of game developers simply dont have skillset to work with audio, graphics APIs and cross platform support.
  - JohnFen 3y ago
    > What you describing is only possible for hobby game developers with extremely small scale of a game or development time of 4-5 years. This is demonstrably untrue. It wasn't all that long ago that there were no game engines available for use, and yet small devs managed to produce a ton of great games anyway. It is true that a small developer is unlikely to be able to produce a grand AAA-scale game in a reasonable time, though. > Majority of game developers simply dont have skillset to work with audio, graphics APIs and cross platform support. Then perhaps they should gain that skillset? I would have assumed that any game dev is well-versed in doing these things regardless, but if not, isn't that a problem?
    - SXX 3y ago
      > This is demonstrably untrue. It wasn't all that long ago that there were no game engines available for use, and yet small devs managed to produce a ton of great games anyway. Audience expectations for production quality of art, animations and overall gameplay has grown accordingly over the years. To become a success game dont have to be best of the best, but it's just cant be much worse than baseline of modern successful games. If there is any indie games that you like and they were released less than 10 years ago please do a little research of how long they actually took to build. There are reasons why one-guy indie game production take 5-7 years. > Then perhaps they should gain that skillset? It's just too many skills and each of them take years of practice. If you take any modern game engine there will be dedicated programmers working on audio, graphics, networking, physics, etc. If you have group of 3 rock start programmer buddies who can work full time for free then you might have some chances, but I doubt you will have any time left to actually make games. Also even with your 3 buddies working for free process of making a game costs money. And if you tell the publishers or investors that your company building their own game engine no one will ever invest money into your project.
      - JohnFen 3y ago
        > If there is any indie games that you like and they were released less than 10 years ago please do a little research of how long they actually took to build. I know a number of successful game developers that are counterexamples to your thesis. > If you take any modern game engine there will be dedicated programmers working on audio, graphics, networking, physics, etc. Right, but that's to make a full engine. I'm arguing that you shouldn't make an engine at all. Yes, if you're aiming at high production values, then you're looking at millions of dollars of investment and a full team. But if you're looking to produce an excellent and successful game, the equation can be entirely different. You don't need top-notch graphics, top-notch sounds, etc. You just need to make a great game, and a great game is defined by the gameplay, not by the window dressing. If you're a one-man shop, it absolutely is major factor in terms of what sort of games you can produce. There's no question about it. But it doesn't mean you can't produce excellent and successful games.
- jayd16 3y ago
  Might as well write a custom language and model editor too. These are all worthwhile endeavors but they are distractions if your goal is the game itself. >you're still creating games. No, every moment you're writing or debugging engine code is a moment you're not working on content.
  - fragmede 3y ago
    The Sapir-Whorf Hypothesis, but for gaming. If the game engine won't let you, say, jump, then the content being generated is limited by that facet of the game engine. So in working on the game engine, and adding the ability to jump, you're unlocking the possibility of new content.
    - jayd16 3y ago
      Or you could pick a modern engine, already have those things and focus on the novel content and not the possibility of content.
- regularjack 3y ago
  This is the same argument as "you don't need a web framework, just build the website". The thing is, by the time the website is done, you have built a framework, and it's worse than all of the existing off-the-shelf frameworks you could've used.

Not to be snide but how many games have you shipped like that? Its really a massive undertaking to build a game engine. Not to mention a distraction as well.

by jayd16 3y ago

Not to be snide but how many games have you shipped like that? Its really a massive undertaking to build a game engine. Not to mention a distraction as well.

JohnFen 3y ago

You don't need to build a game engine, though. Just build a game. Game engines exist basically to be able to reuse code across the development of many games. If you don't worry about writing an engine, and just write games without one, then as a side-effect you'll also be developing the core of an engine that's more specifically optimized for the sort of work that you actually do. Eventually, it may become advantageous to make that into a formal engine, or it may not. Either way, you're still creating games.

SXX 3y ago
What you describing is only possible for hobby game developers with extremely small scale of a game or development time of 4-5 years. Majority of game developers simply dont have skillset to work with audio, graphics APIs and cross platform support.
- JohnFen 3y ago
  > What you describing is only possible for hobby game developers with extremely small scale of a game or development time of 4-5 years. This is demonstrably untrue. It wasn't all that long ago that there were no game engines available for use, and yet small devs managed to produce a ton of great games anyway. It is true that a small developer is unlikely to be able to produce a grand AAA-scale game in a reasonable time, though. > Majority of game developers simply dont have skillset to work with audio, graphics APIs and cross platform support. Then perhaps they should gain that skillset? I would have assumed that any game dev is well-versed in doing these things regardless, but if not, isn't that a problem?
  - SXX 3y ago
    > This is demonstrably untrue. It wasn't all that long ago that there were no game engines available for use, and yet small devs managed to produce a ton of great games anyway. Audience expectations for production quality of art, animations and overall gameplay has grown accordingly over the years. To become a success game dont have to be best of the best, but it's just cant be much worse than baseline of modern successful games. If there is any indie games that you like and they were released less than 10 years ago please do a little research of how long they actually took to build. There are reasons why one-guy indie game production take 5-7 years. > Then perhaps they should gain that skillset? It's just too many skills and each of them take years of practice. If you take any modern game engine there will be dedicated programmers working on audio, graphics, networking, physics, etc. If you have group of 3 rock start programmer buddies who can work full time for free then you might have some chances, but I doubt you will have any time left to actually make games. Also even with your 3 buddies working for free process of making a game costs money. And if you tell the publishers or investors that your company building their own game engine no one will ever invest money into your project.
    - JohnFen 3y ago
      > If there is any indie games that you like and they were released less than 10 years ago please do a little research of how long they actually took to build. I know a number of successful game developers that are counterexamples to your thesis. > If you take any modern game engine there will be dedicated programmers working on audio, graphics, networking, physics, etc. Right, but that's to make a full engine. I'm arguing that you shouldn't make an engine at all. Yes, if you're aiming at high production values, then you're looking at millions of dollars of investment and a full team. But if you're looking to produce an excellent and successful game, the equation can be entirely different. You don't need top-notch graphics, top-notch sounds, etc. You just need to make a great game, and a great game is defined by the gameplay, not by the window dressing. If you're a one-man shop, it absolutely is major factor in terms of what sort of games you can produce. There's no question about it. But it doesn't mean you can't produce excellent and successful games.
jayd16 3y ago
Might as well write a custom language and model editor too. These are all worthwhile endeavors but they are distractions if your goal is the game itself. >you're still creating games. No, every moment you're writing or debugging engine code is a moment you're not working on content.
- fragmede 3y ago
  The Sapir-Whorf Hypothesis, but for gaming. If the game engine won't let you, say, jump, then the content being generated is limited by that facet of the game engine. So in working on the game engine, and adding the ability to jump, you're unlocking the possibility of new content.
  - jayd16 3y ago
    Or you could pick a modern engine, already have those things and focus on the novel content and not the possibility of content.
regularjack 3y ago
This is the same argument as "you don't need a web framework, just build the website". The thing is, by the time the website is done, you have built a framework, and it's worse than all of the existing off-the-shelf frameworks you could've used.