Is there an engine in your heart?

The master once said, "Every game programmer has a engine ". Have you heard of it? Okay, that's what I said. But I believe that every game or graphic program has an engine in its mind. He may never finish it or even start it, but there is always a place in his mind to keep it for this purpose, weaving the wings of his dream there, hoping that one day he could fly out of the original dream that made him feel boring and reach his glorious place. Even if he always feels powerless, even if he knows that it may never be done, he never wants to give up.

Many people think that there are so many engines in the world that there is no need to do it on their own. Indeed, there are many engines that can be found, but few are actually suitable for use. Let's take a look:

1. The gender, Unreal, Cry Engine, Big World, and Gamebryo may be great, but they are too expensive. They are all over knives. Generally, companies cannot afford them, not to mention small teams.

2. for example, the typical Torque and Unity 3D which have developed rapidly over the past two years, Torque is too old, the Code style is obscure, and the performance is average, the only highlight is the seamless indoor and outdoor performance of BSP + Portal + QuadTree, but the problem BSP modeling tool is too difficult to use and cannot provide a good creative experience for art. U3D design is very modern and has good performance. It is very optimistic about it and has the opportunity to try it.

3. Open-source, many open-source engines have their respective highlights, but the open-source problem is that it is likely that no one cares about the open-source engine. Currently, only OGRE is unique. OGRE code is clear and easy to read, with flexible structure and strong scalability. However, flexibility and performance often conflict. The Renderer also follows the original design for a fixed assembly line. The shader always looks like a second-class citizen. It is a bit difficult to adjust as it is completely based on the Design Trend of the programmable assembly line.

How many choices do you have?

I was always thinking about a problem. Why are there so many cool people in China but there are almost no excellent engines. Now I am getting to know something. I think to build an engine, at least meet the following requirements.

1. Personal abilities. Personal ability is the foundation of everything. At least programming languages, data structures, mathematical basics, Windows and 3D APIs, real-time graphics have to go. It is indeed not easy to learn these things well. Many people cannot do it, but there are still many Chinese users who meet this condition. There are still many cool people, but unfortunately there are often not enough people. There are fewer cool people.

2. Ample periods. It may take at least one to two years for the engine to be designed and used. But how many companies can tolerate a one-to-two-year R & D cycle without any output or benefits? Therefore, this condition seems easy to actually be harsh. Most of domestic companies are impetuous in terms of profit pressure. It may be the only chance for large companies to develop engines along with projects. A small company, don't think about it. People point to you for one year. Therefore, we often see many masters on campus, but there are few adult masters, because they lack the environment to grow into masters.

3. With the support of the creation team, if you think that you can build a car by yourself for years, you will be wrong. In the era of programmable pipelines, it is not difficult to create a simple rendering framework to achieve several Amazing effects. The effect will be cool for a few days, adjust the rendering process, and then write a few small shader. Post Effect is much easier, and it is enough to implement Post Effect. Where are GPU Gems, ShaderX, and Crysis. But can you use these cool features for art? How can we arm these features in the creative idea of art? This requires a lot of communication and reflection. More importantly, it is a tool. If you want to build a sound engine, you can make the Content Pipeline intuitive, efficient, and standard, it is not as easy as writing a small plug-in to maximize the productivity of an artist. You have to do too much.

You can think about the number of people who meet these conditions. Of course, this is only necessary but not sufficient. Doing this does not mean you can succeed.

Of course, it is not ruled out that there are some amateurs who are not afraid of difficulties. The Shanzhai engine was developed during the 08-09 period and developed along with our Art Team. Very few features, and the design and performance are not very good. Although the engine is poor, it has been successfully applied to multiple small commercial projects. Recently, it is found that the improvement is more and more difficult, because it is always necessary to consider the compatibility with previous projects, when you need to fear the compromise of history, it may mean you should have a new start.

I don't know when new things can be made or whether they can be made. After a while, I feel that my abilities are too far away, and it may be difficult to learn how to be smart. Learn more first. There were too many delays in the previous paragraph. Please try your best in any case. I would like to tell you that you are not alone. :)

GG GL