The engine is not just an effect, but a program is not just a planning.

In recent work, unreal has discovered some "hidden" things.

For example, test track integration.

For example, UDS: the script is integrated into Visual Studio.

For example, Sync: A Code merge tool.

......

Add the previous swarm and performance4 integration.

 

Sometimes we think that we can surpass it in terms of effect and in some specific links, but if we really want to complete a commercial engine, the height of unreal will be reached, other engines are not that easy to complete.

Because its standard is not just awesome, dazzling, cool, handsome, but high-speed, efficient, and highly integrated development.

When we are still excited about a specific technology, sometimes we may ignore something and what the software is doing.

We always warn ourselves that we are game programmers, so we are not programmers, so we should not use the programmer's set of things to educate us.

We always warn ourselves that we are engine programmers, so we are not programmers, so we should not use the programmer's set of things to educate us.

But now we are increasingly unable to surpass those of programmers.

 

Programmers can analyze the requirements and design the system. I have seen many game programmers who have only one to one planning requirement, two to two, and only a few to achieve three. It is completely reduced to a planning OEM tool. I have also seen many game programmers. The principle I stick to is that you can use it as you plan. Don't struggle ...... I am in charge of fire, I am in charge of death, and planning is completely reduced to program slaves.

Both development models can make excellent projects, which proves how great the Chinese descendants are! Behind these models, there are endless iterations, endless overwrites, and resetting, in other words, none of these projects rely on people who have become accustomed to being brightly lit, full of flesh, and stick to the crazy will under endless pressure.

I have no intention to draw a line between myself and these two types of ideas. In this case, sometimes I feel the hardships of the situation better. In many cases, you don't want to, but you don't have time or energy, no good partener, no patient investor ......

But sometimes I still want to do something new, because the program should be like this.

People are afraid to recognize the dead, and recognize the dead, and simple things will be troublesome.

Yes, it's easy to work with a program. It's easy to plan and work in a program. Why bother yourself.

 

However, after all, the design is not just to reproduce the demand, but to meet the demand, but higher than the demand.

Sometimes it's just because it should have been like this, and there's no other reason.

 

Because no matter the waterfall model or the limit model, it solves only one problem --

There is a conflict between demand uncertainty and procedural stability.

In a simple start, almost all of them are unpredictable.

In an attempt to determine the future, it is not necessary to commit suicide.