Re: Option 1: Alliance stewardship of Boost

From: Vinnie Falco (vinnie.falco_at_[hidden])
Date: 2024-08-06 16:23:20

On Mon, Aug 5, 2024 at 8:14 PM Vinnie Falco <vinnie.falco_at_[hidden]> wrote:

> I would like the opportunity to help it become a many-generational
> institution singularly driven to support library development.
>

All of the Alliance activities are designed to drive up the number of
volunteers and stabilize the project. The new website, a new logo (yes I
know some people hate it), brochures and shirts at CppCon, Boost-related
talks, supporting authors directly through the Staff Engineers program,
continuous integration, better documentation, and IT support. Boost needs
this because there is currently nothing in place to bring in new people to
replace the older ones who depart the project for various reasons. Debates
about what precisely needs to be done are all well and good, but action is
better.

To demonstrate the significance of this problem, I have written a script
which counts the number of unique first names that have committed to any
Boost library repository in a given year. This is what that graph looks
like:

[image: ua.jpg]

The bash script is here:
https://gist.github.com/vinniefalco/81a31b26b8bc9aa371ace8d065eec852

And this is a Google sheet with the output. I manually smoothed 2006
because it was an outlier.
https://docs.google.com/spreadsheets/d/1BENN9T5bzX_aZp0FhlsGW5kcSuoTFm8M7oMJpz3pagk

The Formal Review Process is a key social technology, and it works best
when it has higher participation rates. Growing the number of volunteers or
at least keeping the number from diminishing is important to support this.

2012 marks the transition from SVN to Git I believe. Beman retired in 2018
and passed away in 2020. The COVID epidemic was also in 2020. I am not sure
if this affected the number of unique authors. If these calculations are
correct, it indicates a dwindling supply of authors and by extension
reviewers, review managers, and such. I hope maybe someone can find a big
flaw with my script which invalidates the shape of the graph.

Thanks