Re: Boost Asset Stewardship Review begins 09/03 (Updated)

From: John Maddock (jz.maddock_at_[hidden])
Date: 2024-09-06 12:31:48

On 31/08/2024 00:10, Glen Fernandes via Boost wrote:
> Update: The Boost Foundation proposal mentioned in the original
> announcement has been published.
>
> The Boost Foundation proposal is attached to this email:
> BoostFoundationStewardshipProposal.pdf
>
> A link to the above will also be available in the review commencement
> email that will be posted on September 3rd.
>
> For reference, the review schedule is:
> https://www.boost.org/community/review_schedule.html
>
> The C++ Alliance submitted the following proposal:
> https://cppalliance.org/pdf/Fiscal-Sponsorship-Proposal.pdf
>
> The original announcement email included an attachment of the above:
> https://lists.boost.org/Archives/boost/2024/08/257569.php

First of all thank you Glen for taking on this thankless task!

My review is coming in a bit early and before discussion has really 
taken off, as I'm off on holiday and out of contact from Sunday.

I have no financial connection to either organization, although I do 
know people who are either employed by, or on the board of, both 
organizations.

Boost Foundation

~~~~~~~~~~~~~~

The status quo candidate I guess, on the plus side, they have gamely 
kept the lights on for some time now - for which they deserve enormous 
credit, they have also largely left Boost to it's own devices which has 
mostly been a good thing.

On the negative side they failed to point boost.org to the new website 
even when the community was in favor: something which would have avoided 
this whole issue.  We also have an aging server running obsolete 
software which needs to be actively upgraded, or the lights really will 
go off (including this mailing list).  From what I see in their proposal 
the intention is to replace the mailing list with Discourse, is that 
correct?  I do understand the motivations behind that, but the last time 
that was suggested around here, there was a proper "over my dead body" 
moment.

C++ Alliance

~~~~~~~~~~

The change candidate.  At least somewhat: the Alliance is already 
providing download services, CI, plus engineers to make sure the 
releases actually go out.  On the plus side they have the resources that 
Boost needs - we are a surprisingly rapacious project, something that 
only gets worse the more successful the project is.  They also have a 
clear plan for aging/failing resources, and a clear desire to revitalize 
Boost.  I also like that the Alliance has been actively publishing 
summaries of activities and finances on the mailing list: I'm sure there 
is always more that can be done to improve transparency, but by and 
large it feels like the Alliance has been more pro-active on this front 
than the Foundation.

On the negative side, Vinnie is more from the "move fast and break 
stuff" school of working which may well have upset some. And of course 
untested as custodians of boost.org.  Set against that, I do like the 
formal agreement Vinnie has proposed - more in line with the deal we had 
with SFC - there is a clarity there that is currently lacking from the 
Foundation.  I'm also impressed that Vinnie has responded to criticism 
by actively seeking out new and better ways of organizing things.

Dangers

~~~~~~~

In essence both parties have the same main danger - if they go belly up 
then boost.org goes up to auction.  Everything else can largely be 
replaced, albeit our cloth would have to be cut much tighter, the domain 
name would probably never be recovered though.

Inclusivity

~~~~~~~~

Since the Foundation mentions this in their pitch I though it was worth 
raising.  IMO Boost has always had a strong policy on respectful debate: 
we have been a remarkably politics and personality free forum since 
inception.  The Foundation is none the less correct that we have a 
distinct lack of female authors: this is of course an industry wide 
issue, and one that's super hard to solve for a project composed of 
volunteers.  In short we are dependent on whoever turns up.  And of 
course inclusivity also extends further than gender: to non-native 
English speakers to name but one.  I suspect this issue should be 
treated as orthogonal to whomever controls boost.org, it would none the 
less be interesting to have some concrete suggestions for improving our 
inclusiveness, whether by gender or nationality/language.  In short I 
see this as another facet to increasing participation in general.

Conclusion

~~~~~~~~~

There are strengths and weaknesses on all sides, but as things stand, I 
vote to move control of boost.org's assets to the C++ Alliance.  To me 
it seems like the balance of risks is lower on that side, but also we 
gain the support of a group of people who are not only already committed 
to Boost, but also committed to injecting more energy and involvement in 
the project.

I also strongly hope, that whatever the outcome of this review, all 
parties continue to be involved in Boost in some way: we have a history 
of more than one organization supporting Boost, whether it was Microsoft 
Research sponsoring library development, or via Dave Abraham's Boost 
Consulting.  This always generates tensions when one party has so many 
more hours to devote to the project than anyone else, that volunteers 
struggle to keep up.  We got through things successfully then, and I'm 
sure we will again.

Regards, John Maddock.