On Tue, Aug 1, 2023, 12:21 Hassan Sajjad via SG15 <sg15_at_[hidden]>
wrote:

> I will like to showcase my build-system HMake
> https://github.com/HassanSajjad-302/HMake.
>

It's always great to see experimentation in this space. While this is not
an approach I would personally use, I think having people trying out other
approaches is always good.

HMake however has a flaw and there is no easy solution for it. To fix this,
> I will like to propose a new way to compile source files.
>
> https://github.com/HassanSajjad-302/HMake/wiki/HMake-Flaw-And-Its-Possible-Fixes
>

The problem you highlighted was one of the main points I raised in a recent
talk in C++Now ( https://youtu.be/_LGR0U5Opdg for the video ).

The good news is that in our last meeting in Varna, we actually arrived at
a consensus on how to solve that problem.

We agreed that header units should have a metadata file collocated with
itself to document that it's importable. I'm still a bit behind on adding
the details of that into the draft modules ecosystem TR, but, to
summarize...

The dependency scanner should be able to identify if a header is importable
or not by querying the filesystem without the need to receive the full list
of header units up front.

Daniel