On Boost and modules: second part

From: Ruben Perez (rubenperez038_at_[hidden])
Date: 2024-04-18 13:48:00

Hi all,

Following the recent discussion on Boost and C++20 modules, I've been
performing further investigations on the possible benefits and costs
of offering module support in Boost. I've summarized my findings here:
https://anarthal.github.io/cppblog/modules2

TL;DR:
* While there isn't too much benefit for clean builds, incremental
builds tend to be much faster with modules.
* Right now, modularizing requires an effort for each modularized library.
* Writing CMake files for module consumption won't be trivial.

I'd like to know everyone's opinion on this: do you think these
efforts would make sense? In the mid-term, this will require authors
to review and merge module-related PRs, at the very least. I'm
especially interested in hearing René and Peter's opinions, as
maintainers of B2, Boost.CMake and many other core parts (as well as
other authors').

If the community decides these efforts may be worthwhile, I volunteer
to perform most of the required work. My first step would be to create
a working prototype with a modularized Boost.Url (as a maintained,
separately compiled library, with non-trivial dependencies), together
with CMake tooling, to spot any pitfalls I've missed during my
analysis.

I'd also like to drag any present MSVC developers' attention on this
bug which has nasty consequences on code like Asio:
https://developercommunity.visualstudio.com/t/Using-%5f%5ftry-in-an-inline-function-in-a-h/10186252

Thanks,
Ruben.