Introduction

There are only two kinds of C/C++ Software Build Systems:

1. Good.

2. Not so Good.

3. NO. There is no third kind.

Every software has a build system.

There is never a missing C/C++ Software Build Systems. There is neither a bad C/C++ Software Build Systems. It’s always the case of the above listed two categories. If there is a case without any C/C++ Software Build Systems, then there is no Software at all. By whatever crude — half cooked — incomplete — fragile — temporary — mechanism the Software builds, it still a C/C++ Software Build Systems. Even if it’s not good at doing it, but it still serves the purpose of a C/C++ Software Build Systems — build the Software. Whether good enough to help the developers, or annoyingly bad and remain a painful experience for the developers, a build system would always be there.

The not so good C/C++ Software Build Systems would eventually be replaced by something that is better than the existing C/C++ Software Build Systems. On the contrary, the Good C/C++ Software Build Systems would never undergo any major/visible change. And for most of the developers, such a C/C++ Software Build Systems would be a very opaque system. A good C/C++ Software Build Systems would, on its own, work as an invisible gel that is used to build up the Software stack.

It is not a good idea to learn from a not so good Software build system. And, since a good C/C++ Software Build Systems would evolve to be an invisible/opaque system, it would be a difficult task to learn from such a system.

To cover above scenarios, approach of this book is to take things practically.