read VS Experimental Boost.DI

It was on the mailing list that screens proposals prior to writing papers. Basically it was for making input much easier to work with, like this: https://github.com/ShakaUVM/read

I have thought about submitting a library of mine (https://github.com/ShakaUVM/read) for standardization or inclusion in Boost. Basically it does for input what format did for output - replace the stream system with something more functional.

Using readlib:

That's why I wrote my read library (https://github.com/ShakaUVM/read). You can just do:

5) Make input come from functions rather than call by reference, such as what I did here: https://github.com/ShakaUVM/read

I love: https://github.com/boost-ext/di for dependency injection

https://github.com/boost-ext/di (To detect constructor parameters and inject dependencies without runtime dispatch)

I was looking to try boost.di for some time, looks nice. https://boost-ext.github.io/di/

Things like Dependency Injection/Inversion are a little more cumbersome in C++ but certainly can be done, and I believe there's libraries for that as well (found a couple searching around, like fruit and boost-ext di ).

Dependency injection does not have to be dynamic, it can totally be done at compile time. Boost DI is an example: https://boost-ext.github.io/di/

Funny you mention dependency injection, proposed Boost.DI shall be up for Boost peer review probably in March. https://boost-ext.github.io/di/