On 15/02/2024 18:48, Peter Dimov via Boost wrote:
It's been my impression that the current leadership of the Boost Foundation would like to foster a more inclusive environment for discussions, by way of having an appropriate code of conduct and its
corresponding enforcement.
This, while commendable, is at odds with our traditional culture of frank and merciless technical discussion. So I'm not sure I'd like the new BF moderated forum you seem to propose.
I think, part of why this discussion culture have succeeded in Boost is the mailing lists, which required registration and pre-moderation of the first post. Most of the time, this was enough of a barrier against spammers and trolls. I'm afraid, a more open platform for user communication will remove that barrier and reduce the discussion quality. I think, some moderation will be inevitable, if we want our discussions to be more productive than e.g. Reddit. But I'm happy to be wrong on this.
Some moderation will certainly be necessary. The question is who will be doing the moderating.
I think this is an important discussion, thanks for bringing it. My impression is that many times CoC + enforcement processes don't improve the status quo nor make the community more inclusive (and each one of us usually has a different definition for that word, I'm afraid). I don't think the inclusion level is Boost health's weak point, many people from different parts of the world contribute to this "(con)federated" project, and are motivated by technical excellence opportunities (collaborating, discussing, learning with top C++ experts). Like any other open source project, especially for those not heavily funded, it's hard to maintain, but it's, at least in my case, about the passion of building something useful and keeping it, if not relevant, at least useful. A project that lasts 20+ years is something to be proud of. Boost is also special in its architecture where each library author has a lot freedom, which IMHO is a strong point to achieve long running maintainers. In some sense, it's a distributed effort, with all the advantages and disadvantages of that model.
Just because the Boost Foundation has the magic word Boost in its name doesn't mean we necessarily want to give it any power over the day to day operations of the Boost project. Traditionally, the steering committee had none, and didn't even own anything (because it wasn't a legal entity.)
Hardware and services were provided by people and companies willing to help Boost, for free (e.g. Michael Caisse and Ciere Consulting, but there are many others.)
(We're all familiar with what happened the last time the steering committee tried to interfere by announcing that Boost will be switching to CMake - Rene ragequit and nobody switched to CMake.)
The primary activity of the Boost Foundation is to organize and run C++Now, and running a conference in 2024 does require codes of conduct, banning people deemed to make others feel unsafe, and other similar things. But we the project have never had to conform to these rules.
First of all, I'm really grateful to all that have participated in the Boost Steering Committee and the Boost Foundation for their invaluable effort. I'm grateful to all C++ Alliance members for the same reasons. I am sure that all of them have contributed with their best intention, sacrificing time and money for a good goal, their views don't need to be aligned with my views at all to recognize their merit and effort. Tensions and discussions are part of a vivid and healthy environment. Sometimes having success means that an organizations should evolve, and my view is that organizing and running C++Now, which is a very successful and important C++ conference should not longer be an activity that the entity (legal or not) that wants to steer/advice/improve/push Boost libraries should manage. C++Now is great conference, it should have its own life and management. C++Now currently has little Boost-focused content (I think in 2023 a Boost.MP11 talk was in the schedule, but I don't think there was more Boost content, feel free to correct me). In my humble opinion, Boost libraries need a more focused, less general, technically centered coordination/collaboration entity. I envision more engagement could arise from actual maintainers in the entity, those who write the code and commit the bug fixes, if the mission is nearer Boost libraries' technical aspects and not in procedure and/or conference challenges. That does not mean C++Now and Boost libraries can't cooperate effectively, e.g. having a Boost track in the conference, or having selected talks about Boost libraries... Just like in Boost "We emphasize libraries that work well with the C++ Standard Library", the Boost steering entity should work well with C++ conference sterring entities. But focus is a key aspect for a small community like Boost to take advantage of every ounce of enthusiasm from contributors. Technical aspects are what unite us in this project. I seriously think Boost should part ways with C++Now for the benefit of both. Programmers can contribute to multiple projectsif they want, just like they can be in Boost and the Standard C++ Foundation, Cpp Alliance, CppNow, Meeting C++, Qt Foundation, FSF...
In the old days, the SC would probably just say "thank you Vinnie, very nice, do run the site for us and let the community appoint the moderators, we're very thankful for your eating the substantial costs", but this may not be possible today.
I think your quote from the old days is the Boost way of going forward, let facilitate confidence in those that do the work. Tools for library documentation, the website, the old trac, the regression testing framework and matrix, the incubator, migration to github, etc. happened because we have contributors that are willing to do the work. The community just blessed their effort with confidence and until now, results have been positive. Current website is certainly from another era, many contents are outdated, and... I can live with it. However, the new proposed website is quite ready, it looks good, so let's discuss it to iron those final steps (log in, data collection, whatever) and let's put it in production to improve it. IMHO that's the Boost way. My (long) two cents, Ion