Jump to content
  • Advertisement
DaTueOwner

Anyone who wants to write a little game engine?

Recommended Posts

21 hours ago, Makusik Fedakusik said:

hmph, it's time to become.... sse/avx/neon hacker, gapi hacker, physics hacker, network hacker and other.

Of cource hardware optimisation (that able to give 10x speed-up of computations) is very importent where it can work together with algorithmic optimisation (that able to reduce required calculations to thens and hudreds of thousands  times). For example for collision detection non-chache friendly space partiation algos will be many times fatser than ideally hardware optimized each-to-each  tests. And collision prediction algos is much faster and accurate than any collision detection algos, becouse require recalculate something for collided objects only and at collision time only, instead of every frame tests for each object for collisions detection.

So it again a mathematical/phisical trics, covered by thin layer of hardware optimisation trics where it possible.

Share this post


Link to post
Share on other sites
Advertisement
On 1/18/2019 at 10:12 PM, DaTueOwner said:

It was always my dream to create a little game engine (or rather a graphics engine) for fun

The word "engine" gets thrown around too much these days. And every time someone says it, a bunch of people show up to argue about what an engine is and why you shouldn't make one. (case in point: this topic) It's like they do it on purpose.

My advice to you is to never use the word "engine" if you can help it. You will get a much better discussion that way. In your case, just say that you want to make "a renderer". That should get you more useful responses. And it seems to be closer to what you really want to do.

Share this post


Link to post
Share on other sites
On 1/20/2019 at 4:16 AM, Irusan, son of Arusan said:

No game models this or anything close to it. Graphics is the art of imitating physical reality not the art of reproducing it.

This. And it applies to physics and AI, too. Video games are fundamentally interactive magic shows - "smoke and mirrors." 

Very few games simulate anything you see on the screen with a high degree of physical accuracy. And this is usually intentional - artists and designers tend not to even want physical accuracy. They want the game to look like the image they have in their head of how it should look. They almost always prefer gameplay that is "fun" over gameplay that has accurate physics. They don't want AI that is genuinely "smart," they want AI that feels smart and loses to the player in an interesting way. 

In Doom, why do barrels of radioactive waste explode when you shoot them? Radioactive waste is not inherently explosive. Answer: because it's fun and players love the gameplay opportunities it affords.

Edited by Oberon_Command

Share this post


Link to post
Share on other sites
Quote

Only requirment - no pyton,js or something else that have outdated GC and and memory managment model that not acceptable for realtime software.

This is arguable. Maybe we have two difference definitions on what realtime means, and I'm assuming your definition isn't as nuanced as it should be. 

Quote

Really i could not se anythin but C++ that fit it requitments.

Looking, and seeing are two different acts. The latter implies the answer is in front of you, the former implies you did some prior research before making an assertion ;)

Edited by markypooch

Share this post


Link to post
Share on other sites
2 hours ago, markypooch said:

Maybe we have two difference definitions on what realtime means, and I'm assuming your definition isn't as nuanced as it should be.

Realtime means computation with speed enougt to control process. It lead to other mean - presence of time limit for system response. It lead to requirment of predicability of  executon time for each fragment of code.  GC can not fit it requirment due to unpredicable moment and time required to "stop the word processing". Really it basics of realtime software. Also any scripting or GC language violate other requirments for reliable software - it have to  not use heap reallocation. Becouse only way to warranty presence of required memory is to alocate it on start of system.   

2 hours ago, markypooch said:

Looking, and seeing are two different acts

Pascal is die. Ada and Fortran to rare and have no modern language instruments. C is outdated and have no OOP. And it no any other native OOP languages exists. Also C++ have most advanced instrumemts intendend to build high-level abstraction and  creation of flexible automatic memory and resources managment tools. 

Edited by Fulcrum.013

Share this post


Link to post
Share on other sites
11 hours ago, Oberon_Command said:

Very few games simulate anything you see on the screen with a high degree of physical accuracy. And this is usually intentional - artists and designers tend not to even want physical accuracy.

The reason i want accuracy and robustness in game physics does not contradict your opinion: I want it to avoid things like jittering dead ragdoll bodies, players getting stuck in complex geometry, a box that jitters just because i hold it against a wall with gravity gun, etc. Those things are illusion breakers and the wall of smoke and mirrors becomes obvious much too often. This is no fun.

This just said to point out an aim towards doing things correctly does not go too far, even if it's just games what we are after. It is always profitable towards any goal of game design. It is always possible to tweak a robust physics simulator to do non realistic things if desired, but mostly impossible to tweak a limited simulator to solve harder problems robustly. The same thing applies graphics as well.

Share this post


Link to post
Share on other sites

@JoeJ Are you sure that physics glitches are no fun? People like to share them on the internet and laugh at them so much that sometimes I wonder if developers leave them in games on purpose, just for the sake of guerrilla marketing.

Share this post


Link to post
Share on other sites
1 hour ago, 1024 said:

@JoeJ Are you sure that physics glitches are no fun? People like to share them on the internet and laugh at them so much that sometimes I wonder if developers leave them in games on purpose, just for the sake of guerrilla marketing.

Are you serious? In case, i'll answer: This kind of marketing utilizes the largest failure of games indusrty: Games are so bad that people prefer to watch YT let's plays instead buying and playing themselves. The same people complain about missing innovation in games and they say old games are better than new ones, new games are actually the same than their predecessors just with other content.

This is a problem now when the market can no longer grow. We need to make better games, games that feel new. Progress in tech is just one option to do so, and progress in game tech is usually about increasing realism (even if used for nonrealistic applications).

Of course this applies not to any game, but for the total average it is true i think.

 

EDIT: Recently i've noticed there is a lot of downvoting going on here? On posts that do not deserve this, like the above (-1). This only makes proper discussion hard. I would no longer dare to disagree because i don't want to be suspected to downvote others. Please use arguments instead downvotes. (+1)

Downvotes are intended to mark wrong content. It's not about opinions... maybe in the lounge but not here.

Edited by JoeJ

Share this post


Link to post
Share on other sites
13 hours ago, Fulcrum.013 said:

Pascal is die. Ada and Fortran to rare and have no modern language instruments. C is outdated and have no OOP. And it no any other native OOP languages exists. Also C++ have most advanced instrumemts intendend to build high-level abstraction and  creation of flexible automatic memory and resources managment tools. 

What about rust? The language is really clean and I like it. Also I think you can get away with managed languages like C# just fine if you don't intend to create an AAA tier engine. Example: https://github.com/xenko3d/xenko

Share this post


Link to post
Share on other sites
8 hours ago, JoeJ said:

The reason i want accuracy and robustness in game physics does not contradict your opinion: I want it to avoid things like jittering dead ragdoll bodies, players getting stuck in complex geometry, a box that jitters just because i hold it against a wall with gravity gun, etc. Those things are illusion breakers and the wall of smoke and mirrors becomes obvious much too often. This is no fun.

Oh, for sure. My point was just that most games don't need to focus on accuracy beyond stuff that's illusion-breaking. Very few space games even try to model orbital mechanics, for instance. You don't need an exactly accurate flight model in an arcade flight simulator. Hell, you may not even need to put in exactly accurate integration to get around stuff like jittering ragdoll bodies - just find a way to detect the jittering and then disable the ragdoll code for that body. :P

6 hours ago, JoeJ said:

This kind of marketing utilizes the largest failure of games indusrty: Games are so bad that people prefer to watch YT let's plays instead buying and playing themselves. The same people complain about missing innovation in games and they say old games are better than new ones, new games are actually the same than their predecessors just with other content.

I don't know. I had a lot of fun messing around in Goat Simulator, but maybe that's the exception that proves the rule here.

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

GameDev.net is your game development community. Create an account for your GameDev Portfolio and participate in the largest developer community in the games industry.

Sign me up!