On 1. Aug 2024, at 14:12, Andrey Semashev via Boost wrote:

On 7/31/24 23:11, Kristen Shaker via Boost wrote:

...

Here are what we believe to be the available options.

1.

The C++ Alliance assumes control of the Boost assets, including the boost.org domain name. The Boost Foundation becomes uninvolved in any decisions related to the Boost Libraries. 2.

The Boost Foundation continues to be the stewards of the boost.org domain name and related assets. New assets that are meant to be associated with the Boost Libraries are transferred to the Boost Foundation. In any matters related to the Boost Libraries, the Board will abide by any decisions made by the developers but will no longer vote themselves on issues as they relate to the Boost Libraries unless there truly is no clear consensus or path forward.

I'm probably not going to be helpful, but I don't really like either of the options, with the first one being slightly less preferable.

Although the input from The C++ Alliance members in many areas of Boost is undeniable, the organization is apparently being run by a single individual. I haven't met Vinnie, so I can only judge about his personality based on his posts to this list and a small amount of personal correspondence. My impression is that he is devoted to Boost, but rather emotional and ambitious, and that may sometimes cloud his judgment. It is possible that his interests won't always be aligned with Boost, and I cannot be sure that his devotion to Boost won't change in the future or that he won't try to transform Boost into something that is not accepted by the wider Boost community. I'm sorry if this sounds like a personal stab at Vinnie, but it really isn't. It is not my intention to offend anyone, but given the structure of The C++ Alliance funding, personalities need to be considered.

Regarding The C++ Alliance organization, its mission statement (https://cppalliance.org/#mission) doesn't even mention Boost. In fact, it focuses on C++ advancement in general and is closer to the Beman Project in spirit. Boost Foundation mission statement (https://sites.google.com/boost.org/boost-foundation/home#h.rszdmunawmm3), I feel, is more aligned specifically with Boost, as it focuses on library development and peer review process, which are effectively what Boost is.

So, in short, I don't like the option 1 because I don't fully trust The C++ Alliance to focus on Boost in the long term. Giving it full exclusive control over key Boost infrastructure elements seems like too much power in one's hands.

Same here, +1 for option 2. I rather have inefficient and slow progress, considering the alternative. Res publica.

On 8/1/24 16:16, Vinnie Falco wrote:

On Thu, Aug 1, 2024 at 5:12 AM Andrey Semashev via Boost wrote:

...

I cannot be sure that his devotion to Boost won't change in the future or that he won't try to transform Boost into something that is not accepted by the wider Boost community.

The Alliance would be administering:

* The boost.org domain * Related cloud services

The Alliance cannot "transform Boost" with this any more than the Foundation could. We do not control the GitHub organization or the library repositories. The new website has been licensed under the BSL and donated. The idea that control of the domain equates to control over Boost is not a serious one.

I disagree. The list above is incomplete, there are also the logo (with the currently controversial terms of use), the website, social network accounts, the mailing list and all related infrastructure. You yourself admitted it is becoming increasingly difficult to pick people who are not affiliated with The C++ Alliance to fill important roles, such as review managers and reviewers. Given that you alone are funding the Alliance, I'd say you have a pretty strong leverage on it, and by extension on Boost.

...

Regarding The C++ Alliance organization, its mission statement (https://cppalliance.org/#mission) doesn't even mention Boost.

Mission statements don't really mean much. They aren't legally enforceable and they can be changed at any time. The behavior over time is more reflective of intent than verbal postures. We could update the Alliance mission statement if you want...

Mission statement is a declaration of intended goals of the organization. I assume it should be taken into account in the decision making process within the organization. And while not legally enforceable externally, it may be enforceable internally, if internal regulations are defined accordingly. In any case, since I'm not a member of The C++ Alliance, it doesn't really matter what I want. Mission statement is what you, the Alliance, want. I'm just comparing the declared goals of the two orgs.

The list above is incomplete, there are also the logo (with the currently controversial terms of use), the website, social network accounts, the mailing list and all related infrastructure. You yourself admitted it is becoming increasingly difficult to pick people who are not affiliated with The C++ Alliance to fill important roles, such as review managers and reviewers. Given that you alone are funding the Alliance, I'd say you have a pretty strong leverage on it, and by extension on Boost.

In these examples, the C++ Alliance is not taking control of anything. It controls the things it creates and maintains so they can continue to exist. And even make things open so other people can continue without them if they ever want to. If person A is providing X, even if you share everything you have done so far, you will always have control over whether you can keep providing X. That's not person A having control over something that exists. It's person A having control over something that might exist. If even this kind of "control" (more like self-determination) is to be avoided, the only solution is for everyone to be prohibited from providing anything because they might one day stop providing it. Unless the point is not about control and more about not wanting to become dependent on person A. But in that case, the solution is to actively find other people who will contribute. The solution is not to stop people from contributing just to avoid a scenario where they might stop contributing one day.

On Thu, Aug 1, 2024 at 7:08 AM Boris Kolpackov via Boost wrote:

3. Nominate someone thoughtful and respected by both camps (like Peter Dimov) as a final arbiter in the disputes between the Alliance and the Foundation when it comes to the Boost business.

Well, it is a bit late for that :) The dispute is resolved with Kristen's latest announcement, which leaves this important decision to the community as it should be. Thanks

On Thu, Aug 1, 2024 at 12:31 PM Boris Kolpackov via Boost wrote:

Vinnie Falco writes:

...

On Thu, Aug 1, 2024 at 7:08 AM Boris Kolpackov via Boost wrote:

...

3. Nominate someone thoughtful and respected by both camps (like Peter Dimov) as a final arbiter in the disputes between the Alliance and the Foundation when it comes to the Boost business.

Well, it is a bit late for that :) The dispute is resolved with Kristen's latest announcement, which leaves this important decision to the community as it should be.

Let me quote what I believe is the relevant part of Kristen's email:

Kristen Shaker via Boost writes:

...

Accordingly, we feel the most appropriate thing to do at this juncture would be to let the developers make a decision on how they would like to proceed regarding what level of ownership the C++ Alliance should have on Boost Library assets.

Here are what we believe to be the available options.

1. The C++ Alliance assumes control of the Boost assets, including the boost.org domain name. The Boost Foundation becomes uninvolved in any decisions related to the Boost Libraries.

2. The Boost Foundation continues to be the stewards of the boost.org domain name and related assets. New assets that are meant to be associated with the Boost Libraries are transferred to the Boost Foundation. In any matters related to the Boost Libraries, the Board will abide by any decisions made by the developers but will no longer vote themselves on issues as they relate to the Boost Libraries unless there truly is no clear consensus or path forward.

Maybe I am reading it wrong, but this essentially says two things:

1. Boost developers should decide how to resolve the impasse.

2. It proposes two possible ways to resolve it.

One important clarification.. It gives three ways to resolve it. But, I think, incorrectly discards the third one. -- -- René Ferdinand Rivera Morell -- Don't Assume Anything -- No Supone Nada -- Robot Dreams - http://robot-dreams.net

Vinnie Falco writes:

Maybe a miscommunication on my part. What I meant is that the dispute between Alliance and Foundation is resolved (because the community is now deliberating).

I don't see how it is resolved. The foundation essentially asked the Boost developers to choose who will control the Boost assets going forward, the Alliance (option 1) or the Foundation (option 2). In particular, they explicitly said that if option 1 is chosen, then the Foundation will walk away from everything Boost. Again, here is the relevant part from Kristen's email: Kristen Shaker via Boost writes:

Here are what we believe to be the available options.

1. The C++ Alliance assumes control of the Boost assets, including the boost.org domain name. The Boost Foundation becomes uninvolved in any decisions related to the Boost Libraries.

We may have a path to resolving this dispute but the dispute is definitely not resolved. I also hope everyone realizes that this is an existential moment for Boost. If option 1 is chosen and some time down the line the Alliance runs out of money or decides to shift their focus elsewhere, there will be no one to catch the balls in the air, so to speak. It will be very hard for a project to go from burning $500K a quarter to zero without imploding.

On Thu, Aug 1, 2024 at 12:59 PM John Maddock via Boost wrote:

On 01/08/2024 13:12, Andrey Semashev via Boost wrote:

...

...

Here are what we believe to be the available options.

1.

The C++ Alliance assumes control of the Boost assets, including the boost.org domain name. The Boost Foundation becomes uninvolved in any decisions related to the Boost Libraries. 2.

The Boost Foundation continues to be the stewards of the boost.org domain name and related assets. New assets that are meant to be associated with the Boost Libraries are transferred to the Boost Foundation. In any matters related to the Boost Libraries, the Board will abide by any decisions made by the developers but will no longer vote themselves on issues as they relate to the Boost Libraries unless there truly is no clear consensus or path forward. I'm probably not going to be helpful, but I don't really like either of

On 7/31/24 23:11, Kristen Shaker via Boost wrote: the options, with the first one being slightly less preferable.

Exactly, it's sort of like asking a panel of folks who don't know each other (the developers) to conduct a job interview with two applicants they also don't know. Expect an entirely random answer!

Interesting analogy and perspective. That suggests that a review process, as suggested elsewhere, for the developers to make a choice is the way to go. Where it would allow the applicants to promote their vita. -- -- René Ferdinand Rivera Morell -- Don't Assume Anything -- No Supone Nada -- Robot Dreams - http://robot-dreams.net

El 02/08/2024 a las 0:53, Ion Gaztañaga via Boost escribió:

El 01/08/2024 a las 20:07, Joaquin M López Muñoz via Boost escribió:

...

El 31/07/2024 a las 22:11, Kristen Shaker via Boost escribió:

...

There was some discussion about the assets being held in a third party non profit controlled by a subset of the Boost Developers. However, given that there are already many Boost Developers on the Boost Foundation Board of Directors, we don’t see this as a meaningful deviation from the status quo.

I think that your argument that a potential third party controlled by a subset of the Boost Developers is not meaningfully different to the Boost Foundation is incorrect:

[...]

Hi Joaquín,

Just to try to clarify a bit the "third option" that John and you are mentioning (and think some other have also suggested), as I understand it:

- A new non-profit entity

- Controlled by a subset of active and prominent Boost Developers (e.g. those should be the overwhelming majority of the board)

- An entity whose primary (unique?) purpose is to support the Boost libraries (no additional projects, no conferences, etc.).

Is this an accurate description of the "third option"? If so, I agree that this is a substantial deviation from the status quo. If other boost developers find this meaningfully different, then this option should definitely be available.

Hi Ion, Yes, this is a good description of how this third option could look like, and yes, the Boost Foundation does not currently look like that. Joaquin M Lopez Munoz

On August 1, 2024 11:15:29 PM Alan de Freitas via Boost wrote:

Regarding option (1), if I understand correctly, the C++ Alliance didn't ask to control existing assets.

They did try to buy the boost.org domain.

On 8/1/24 3:02 PM, Alan de Freitas via Boost wrote:

The Boost Foundation cannot transfer control of the domain to the C++ Alliance because they don't own it either > There's no point in voting on whether they can transfer ownership of something that is not theirs.

Right. The domain name "boost.org" is owned by the Beman Dawes estate and can only be transferred by the executor of that estate. The other asset of interest might be the github repositories. At least in the case of safe_numerics, that repository is "owned" by me - Robert Ramey. I don't have any intention of transfering that ownership to anyone else. I can't speak for the other repositories. Github doesn't reveal who owns them to anyone other than the owner. Finally there is the boost.org "umbrella" repository. I believe that this is "owned" by Glen Fernandez. I put "owned" in quotes as it's unclear to me what the owners property rights in this case. Seems to me that this is beyond the influence of both the Boost Foundation as well as the C++ Alliance. The only thing that might change things is that C++ Alliance and Boost Foundation attempt to purchase the ownership of the domain boost.org from the Dawes estate. It's not really possible to know what would happen then. It might spark a "bidding" war as both parties have significant assets and apparently interest. The "winner" would be the Dawes estate who would receive a payment commensurate with their perceived value of the domain name. Which would be a fine outcome as far as I'm concerned. Beman was a friend of mine. He (along with David Abrahams) worked his butt off for years making Boost a success. I see no better result that his family receive some recompense for his unselfish efforts and contributions to C++. I do now believe that without these two gentleman, C++ would not have survived into the 21st Century. Boost accomplished it's original goal in 2011 when large parts of Boost were incorporated into the C++ standard. It hasn't been very relevant after that. Basically, Boost was so successful, that it worked itself out business - as successful organizations to. Of course these organizations don't die immediately and no one want's to kill them off due to sentimental reasons. Sometimes they can re-purposed and continue on as version 2.0. Jon Kalb took on this task in 2012? and managed to get another 12 years out the franchise and build the successful C++Con. (For this he was rewarded by getting unceremoniously pushed out of Boost by the board of directors of the Boost Foundation due to some code of conduct BS. Shameful and cowardly behavior by members of the board. But I digress. Here we are stuck again. Something's got to give. The status of the library repositories probably won't change much. C++ Alliance will acquire the domain boost.org and develop the website, (I'm not sure why - oh well) or not in which case Boost Foundation will do it (?). Regardless, I see the purpose and goals of Boost significantly changing or Boost dying. Consider this my modest tribute to a life well lived. Robert Ramey PS. The Beman project is living in the past. I predict that it will be nothing but a zombie website within a year. RR

On 01/08/2024 23:42, Robert Ramey via Boost wrote:

Regardless, I see the purpose and goals of Boost significantly changing or Boost dying.  Consider this my modest tribute to a life well lived.

The way I'd look at it Robert is life is change, and something changing significantly means it ain't dead yet. I got some negative comments for saying this last time, but I'll say it again - the proportion of the active users here who personally knew the original set of active users has become a minority. That is a sign of vitality - we are being replaced. I, like you, have had considerable interactions with both sets over the years. Both sets have positives, but both have negatives. Quite a few appear to be yearning for a third option which isn't either of the current two on the menu before us. Unfortunately, politics rarely presents ideal options. You get two options most people don't much like, and the choices are A, or B, or vote for neither. The C++ Alliance choice is very much a put your eggs in one basket in exchange for proactive funding of dev work until the money runs out. The Boost Foundation choice is to continue the benign and sometimes not so benign neglect of dev work and the developers themselves. So long as the source code remains Boost licensed, many will feel it doesn't matter which path is chosen. A fork and a new website could always be done if whomever owns the supporting IP got into trouble. So I suspect most will choose to sit on the fence and see how things play out. The Alliance choice thanks to the direct funding of devs will always have far more energy and bodies to apply to problems. They will choose the problems which match their choice of direction however. There will be an actual direction now for better or for worse. I honestly don't know which is better. One has strength and vitality but it's brittle, the other is continue to meander aimlessly. I suspect the former has less chance of "Boost dying" in the short term, but who knows longer term. I will say this: the future is determined by the young and energetic, and that's ain't us no more. So I guess whatever they think is best. Niall

The C++ Alliance choice is very much a put your eggs in one basket in exchange for proactive funding of dev work until the money runs out. The Boost Foundation choice is to continue the benign and sometimes not so benign neglect of dev work and the developers themselves. That and Am 02.08.24 um 08:03 schrieb Boris Kolpackov via Boost: I also hope everyone realizes that this is an existential moment for Boost. If option 1 is chosen and some time down the line the Alliance runs out of money or decides to shift their focus elsewhere, there will be no one to catch the balls in the air, so to speak. It will be very hard for a project to go from burning $500K a quarter to zero without imploding. My concern is also the apparent shift from a low-cost, low-maintenance (website?) solution to something seemingly requiring vast amounts of money and effort. As far as I understood the old website was more or less static HTML while the new one requires a considerable backend to provide dynamic services. That might not be sustainably long-term or in case someone (in this case

Am 02.08.24 um 09:01 schrieb Niall Douglas via Boost: the C++ Alliance) steps down and someone else has to volunteer to take over. But given the BSL license of the website I'm optimistic that we could extract a static version of the new website (without the user interaction parts) to have it run in a low-cost environment again. Especially the announcement of essentially: "If C++ Alliance takes over (some resources) the Boost foundation will become fully uninvolved/unrelated to Boost" That doesn't sound right. Am 02.08.24 um 00:02 schrieb Alan de Freitas via Boost:

...

...

Regarding option (1), if I understand correctly, the C++ Alliance didn't ask to control existing assets. They did try to buy the boost.org domain. First, and most importantly, that's not what's being voted on. The Boost Foundation cannot transfer control of the domain to the C++ Alliance because they don't own it either. There's no point in voting on whether they can transfer ownership of something that is not theirs.

3) How does it plan to migrate the mailing lists to a server that its not in imminent danger of falling over? That would be the main question. Is it "easy" to transfer the mailing

As far as I followed the discussion about the domain there seem to exist ownership issues. I don't really know the (legal) implications but it sounds like it needs resolving via buying the domain from the current owner and the Boost Foundation as the current "volunteer" related to things like that didn't do that over the last years so someone else stepped up. I would prefer if the domain is ultimately owned by the Boost Foundation as the inactivity related to those matters at least shows that nothing bad will happen. I expect the same (or even improvements) when the Alliance owns it but I do understand the sentiments against having essentially a single person controlling it and what will happen to the domain once the Alliance shifts its interest after having bought it for a large amount of money. What other assets are being talked about here? Maybe: Twitter account(s): I don't mind the Alliance controlling them as in the past they seemed to be the only one with any activity there and seemingly did a good job. I would prefer a shared ownership though such that in case of shifted interest someone else in the Boost community can take over without the accounts getting abandoned by accident. Mailing list: list servers? I don't see any large danger here even if that was controlled by a single person, if that is trusted now. In case that person looses interest someone else can take over or the current status-quo can be re-established. But similar to the domain this seems to be something requiring action and someone has to do that. Download servers: This sounds similar to the mailing list: Immediate action was required, the Alliance did that and in the future we can possibly point download links of (at least new) releases to another location, e.g. the current one even if that isn't optimal and if it was barely working. So I don't think this is highly riskily either. I do like the idea of Joaquin about a new entity consisting of Boost developers to primarily/solely support Boost but I don't think this is viable as Open-Source volunteer work will always suffer from financial issues. But I hope I can be proven wrong here or I'm misunderstanding the implications. So bottom line: If the Boost foundation can and will take appropriate steps to solve the most important issues (that the Alliance is currently solving or intending to solve) I prefer they do so. Otherwise I'm carefully optimistic that things work out with the Alliance going forward. But I, like most devs, don't really want to get into "politics". So I just hope for things to work in the end. Alex

On Fri, Aug 2, 2024 at 1:42 AM Alexander Grund via Boost wrote:

My concern is also the apparent shift from a low-cost, low-maintenance (website?) solution to something seemingly requiring vast amounts of money and effort. As far as I understood the old website was more or less static HTML while the new one requires a considerable backend to provide dynamic services.

The old website is just one service provided by the "wowbagger" server which currently hosts much of the infrastructure required to support Boost. There is also the mailing list, currently using mailman2 which is no longer supported. There are various scripts and services such as the commit-bot which updates the submodules in the superproject as needed, scripts to perform builds on demand as the superproject changes and publish the rendered documentation for master and develop branches to the website, and more.

That might not be sustainably long-term or in case someone (in this case the C++ Alliance) steps down and someone else has to volunteer to take over.

This is a reasonable concern but I think that the fears might be overblown. In the unlikely event that the C++ Alliance needs to focus its efforts elsewhere or cease operations, we would support a graceful transition for the community. The Alliance consists mostly of Boost volunteers just like everyone else except that they are compensated. They have their own libraries which they author and maintain. Matt Borland for example, was already a long-time Boost contributor before becoming an Alliance Staff Engineer. Our people are Boost volunteers first, employees second. Ask anyone who works with us, such as Marshall Clow, for example, who was our first hire. The idea that the Alliance is controlled by one person who can simply walk away and leave Boost worse off, is not a serious one. I for one, have several libraries in Boost and several more in development. And there is no desire for me to abandon my work with all the years I have put into it. I do have concerns about the long-term sustainability of the Boost project, because there is no well-defined plan or visible efforts made to replenish our diminishing ranks with new, talented engineers. This is something I am trying to change, even in the face of some rather vocal opposition to my methods (such as a trademarked logo).

I would prefer if the domain is ultimately owned by the Boost Foundation as the inactivity related to those matters at least shows that nothing bad will happen.

The boost.org domain expired in 2022, becoming unavailable for almost two days. If Beman Dawes' surviving spouse had passed away at any time before the Alliance made its offer to the estate, the boost.org domain would go to auction in 2025.

I expect the same (or even improvements) when the Alliance owns it but I do understand the sentiments against having essentially a single person controlling it and what will happen to the domain once the Alliance shifts its interest after having bought it for a large amount of money.

Our offer to the estate was for a total of $35,000 plus legal fees. The sum is broken down into $5,000 to the executor, and $30,000 to establish the "Beman Scholarship." This offer still stands of course. It is not "a large amount of money." The boost.io domain, on the other hand... that was rather expensive ($135,000). It is looking more and more like we will not be needing it, thankfully.

I would prefer a shared ownership though such that in case of shifted interest ... In case that person looses interest

This seems to be a common, reasonable fear regarding Alliance stewardship of shared resources. Given that Boost's current governance structure has enabled stagnation. I think this is a problem worth solving. I am already putting together a long-term plan for the continued operation of the Alliance when I am no longer present. We could also figure out some kind of backup plan which developers can utilize if Alliance governance becomes unsuited for the project. Yes I think this needs to be addressed. It will be easier to attract new people to the project if it is seen that it has a durable foundation. Thanks

On Fri, Aug 2, 2024 at 5:22 PM Vinnie Falco via Boost wrote:

The Services cost about $13,000 a year to run

Hi, Louis here. I manage the operations at the C++ Alliance. I’m writing to provide more details on the costs for hosting the new website. This is a snapshot of the costs for the month of June for all the services running boost.io: Google Cloud Platform $580 Kubernetes Engine $357 Compute Engine (Databases) $100 Compute Engine (mailman3 servers) $116 Cloud Memorystore for Redis $40 Networking $3 Artifact Registry $3 Cloud Storage Other $50 Amazon S3 Buckets $60 Fastly Website Caching $35 Mailgun service $1,344 Total for June To summarize, the annual projected cost for hosting the new Boost website is $16,128 per year. The board has tasked me with implementing a fundraising plan to achieve financial sustainability for the Boost Software Commons, to replace the loss of Foundation conference revenue and to ensure that Boost is not dependent on Vinnie’s donations for its infrastructure. To this end I have created a Patreon account. Donations are disabled for now, until the Boost Software Commons establishes the infrastructure to receive payments. https://patreon.com/boostorg Thank you, Louis Tatta The C++ Alliance