Boost or not Boost, that is the question?

Question

C++ / C#

How to understand when to use boost and when not?
I am writing a cross-platform network application on the plus side for file synchronization. I just can't decide whether to write in "pure C++" or using boost.
If I write in pure pluses, then I have to use a lot of pure C functions, for example, for the same sockets or operations with lists of files in a folder, and I get a mess of 2 languages. But the project is going to fly. In this case, the thought is already flashing in general to write in pure C. But here you need to write your own bicycles such as timers, a config parser, etc.
If I start implementing the same project with a boost, then everything looks short and in the style of pluses with all the necessary things, but the compilation time of the application increases and you still need to carry boost libraries with you.

Answer 1

Precompiled headers , come on! And everything is going to fly.
Well, or for very difficult cases, you can use Ninja .
And write clearly on the boost.

Answer 2

If you are writing something cross-platform where there will be a lot of work with the network, then the boost is indispensable. If you need to use a couple of small functions from the boost (and the functionality will not be expanded in the future), then you can write it yourself. True, you probably also have to work with the file system ... So most likely you need a boost.
You can also consider other libraries, which, by the way, are not so few.