Do you want to develop a game or an engine? At some point you will have to use a third party library like Direct X, so is it cheating? No, you relatively have no choice if you care about compatibility. Learning how to develop with third party engines is difficult enough and remember you are just one person, there's huge teams developing engines and one person can not get it to the same level of competition and as a learning example is relatively pointless until you know the basics of how a decent game goes together, if you want a challenge take a crack with Unity..

To get it up to a decent graphical standard, chuck the default shaders in the bin and start from scratch with shader lab. Work on adding some advanced features like sorting out the Post AA in DR, things like SMAA shaders. Dump all the post effects and try implementing DSSDO, pristine mostion blur, add other lacking things like a Voxel based realtime GI.. Build your game, sort out all the challenges in AI, controllers, physics, shadows and performance.. Create your own LOD / Culling system is also a good idea.. Come back in a fair few years when you're done. (That's if you have experience and the knowledge to know how to put it all together)..

There is so much to learn about using engines and adding to them, that alone can take many years to master. Never mind trying to build your own engine.! There just isn't enough time, I own a company with a team of 5 very experienced devs and we still don't have enough time and we certainly aren't trying to re-invent the wheel.

The reason I don't think using an external library is cheating is for the same reason I don't believe buying a loaf of bread rather than making it from scratch is cheating.

Again, you could make a hammer from scratch or you can buy one from the store, or borrow a neighbor's.

I don't intend to program the low-level stuff every time I make a game, I just want to have an idea about how it's done, and that's the point where I get frustrated with myself. I am going to use libraries, engines, and just focus on the game, but I would like to make my own mini-engine (not as powerful Unreal Engine) because i find joy in programming and would love to be recruited in a professional company. Reinventing the wheel might be counter-intuitive, but it might also be necessary learning progress, the same way a doctor will study anatomy hands-on instead of just using read material.

Okay, that is perfectly fine and reasonable. It seems that you envision a massive amount of coding far beyond the ability of one programmer - meaning that it typically demands a team - in order to make the usual game engine of professional caliber. Game developers who have looked at it objectively realize that even the smaller game engines took a team at least several years to create it. As long as you have a healthy awareness of your limitations with a comprehensive game engine, then you should be just fine with that plan - a good one at that.

Remember not to spread yourself too thin. I feel that it is better to excel in 2 or 3 major areas instead of dipping into many, only my view.

There is so much to learn about using engines and adding to them, that alone can take many years to master. Never mind trying to build your own engine.! There just isn't enough time, I own a company with a team of 5 very experienced devs and we still don't have enough time and we certainly aren't trying to re-invent the wheel.

The team of which I belong has put many many many hours over the last several years mostly with existing libraries and industry standard workflow pipeline. Guess what? After years we are finally seeing the light at the end of the tunnel and a release of a game in the conceivable future! ... and we are a very skilled TEAM working our... uh... tails off, so a similar situation to yours in many ways.

There is so much to learn about using engines and adding to them, that alone can take many years to master. Never mind trying to build your own engine.! There just isn't enough time, I own a company with a team of 5 very experienced devs and we still don't have enough time and we certainly aren't trying to re-invent the wheel.

The team of which I belong has put many many many hours over the last several years mostly with existing libraries and industry standard workflow pipeline. Guess what? After years we are finally seeing the light at the end of the tunnel and a release of a game in the conceivable future! ... and we are a very skilled TEAM working our... uh... tails off, so a similar situation to yours in many ways.

Very much so, I don't think people realise what's involved until you're at least a fair way in .. (Obviously dependant on what you're trying to achieve), if it's a high fidelity modern 3D game then there's more than enough challenge for three years and a team of 5 just making it. Never mind anything else, sure we are trying to improve our feature set to stay competitive. But that's about it!.. We have a half built engine which we are releasing in 2015, for no more reason than continued revenue after we have finished our game..

I wish you the best of luck with your game and a happy new year.

I wish you the best of luck with your game and a happy new year.

Thank you! ... you too!