On Mon, Mar 11, 2024 at 8:11 PM Robert Ramey via Boost wrote:

On 3/11/24 1:22 PM, Louis Tatta via Boost wrote:

...

On Mon, Mar 11, 2024 at 12:23 PM David Sankel wrote:

...

Our monthly Boost Foundation meeting is this Wednesday at 3pm Eastern (12pm Pacific).

...

Second, as the agenda for the Boost Foundation board meeting occurring on March 13th includes a discussion of the new website,

...

Louis Tatta CEO, The C++ Alliance, Inc.

_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

Is there any reason that the new web site could not be subjected to a process similar to the formal process for new libraries? This would help build a consensus around functionality, design, etc. I just don't see any other way to be sure that all considerations are accounted for.

Having the unique experience of designing and implementing the past two Boost web sites. Designing and implementing a website doesn't lend itself to the same kind of cut and dry accept/reject process that is the formal review process. Mainly because most of the aspects are not technical but subjective. Hence I don't believe such a process would accomplish anything other than some people like the site while others do not. Having said that though.. I think the new site design has undergone a looser version of a formal review process where the question is not "should this be accepted?" but instead "what needs to change so that it's usable?". The goal is different. But the process isn't that different. And more importantly, the process does not end for the site. How to make the site usable will continue to be asked, investigated, and changes implemented. -- -- René Ferdinand Rivera Morell -- Don't Assume Anything -- No Supone Nada -- Robot Dreams - http://robot-dreams.net

I would like something along those lines. That is, I would like a review of functionality, not aesthetics. Re-reading What Ion wrote, I realize I'm not up for a twice-yearly review, so much as one-time input to the web developers on users' needs. Zach On Tue, Mar 12, 2024 at 2:41 AM Ion Gaztañaga via Boost wrote:

El 12/03/2024 a las 2:39, René Ferdinand Rivera Morell via Boost escribió:

...

...

Is there any reason that the new web site could not be subjected to a process similar to the formal process for new libraries? This would help build a consensus around functionality, design, etc. I just don't see any other way to be sure that all considerations are accounted for.

The main difference IMHO is that the mailing list is full of C++ experts that are probably terrible at design and usability ;-) Just look the aesthetics of our templated or preprocessor metaprogramming code!

Now seriously, not a formal review, but I think it's important that main issues (https://github.com/boostorg/website-v2/issues and https://github.com/boostorg/website-v2-docs/issues) could be discussed and monitored periodically (say, every 6 months) in the ML. Would be cpp.al folks fine with this approach?

Best,

Ion

_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

Now seriously, not a formal review, but I think it's important that main issues (https://github.com/boostorg/website-v2/issues and https://github.com/boostorg/website-v2-docs/issues) could be discussed and monitored periodically (say, every 6 months) in the ML. Would be cpp.al folks fine with this approach?

I don't think that's a bad idea, but unless things change I don't think it provides a lot of additional value over what we have now, which is already everything being constantly monitored and people being able to complain and open issues about anything at any time.

On 3/12/24 2:09 PM, Ion Gaztañaga via Boost wrote:

El 12/03/2024 a las 17:24, Alan de Freitas via Boost escribió:

Yes, I just wanted to propose some kind of "state of the website" review once a while from the community. It could be every 12, 24 or 36 months.

The difference with a library is that the website is representing somehow the whole federated project and it seems that some collaborators here might be worried about how the website project will go in the long run.

But yes, if issues are fixed (the website has now 100+ issues), and we have few complaints in the ML, then there is no need to review anything.

I don't think that the whole value of the Boost Review process is being appreciated here. It's unique as far as I'm concerned. It focuses everyone's comments to a short period where they can all be discussed together and the best decision can be made for the the component's inclusion/exclusion. To just say that we'll make the switch and people can bring up issues as they occur is not the same thing. Enough of the web site exists so that it's clear what it would be like. The authors of this work deserve a yes or no answer as to whether it's going to be accepted or not. That is the purpose of the Boost Review. It's an onerous process. Many library authors have declined to submit themselves to it. It is a lot of work. But I think on balance the Boost review process is the secret of Boosts success. For something as central to Boost as this web site, the review process should be applied. I'm asking the Board of Directors to impose this requirement if its not already clear that it's necessary. Robert Ramey

On Tue, Mar 12, 2024 at 2:28 PM Robert Ramey via Boost wrote:

The authors of this work deserve a yes or no answer as to whether it's going to be accepted or not. That is the purpose of the Boost Review.

Thinking about this a bit more, I have some clarity into some of my recent uncomfortable feelings. A C++ library goes through the review process, and whether it passes or fails the library is still usable. The author can derive value from the library even if rejected, by using it in their own projects and encouraging others to use it and form a following. A new Boost website however, is different. If the website is rejected, then most if not all of the work is for nothing. It is a total loss or close to it. We didn't build a new website on a whim, there was a discussion a few years ago about the need for a new site. There are stakeholders from the Boost community that have worked with us. Although the website was not designed by committee (if for no other reason that such a process makes failure likely), we did get community engagement once there was something worth looking at. Unlike a C++ library, a new Boost website is really only usable for becoming a Boost website. The problem with a straight Boost Formal Review applied to the website, is that a rejection would likely discourage anyone from making the effort again in the first place. Case in point, if I was told at the beginning that our only option would be to publish our website on a different domain while David either co-opts our work himself to place on boost.org, or devalues our labor and puts up his own new website on boost.org, we would have never done the work in the first place. The new website reflects my love the Boost Libraries, and in my opinion there needs to be an implicit understanding that if someone takes on the enormous risk of building a new Boost website, delivers on that promise, and demonstrates their understanding and adoption of Boost culture, we want to encourage and celebrate this. As the website (unlike a library) cannot be repurposed, a rejection would not only be an immense personal loss for me but represent a significant loss for Boost at a time when we need all the victories we can achieve. I would ask everyone who loves Boost to rally and get behind this effort to publish this modern website, to show the onlookers and new potential community members that we are capable of supporting this high-risk / high-reward contribution. The alternative is too dismal to contemplate. Thanks

On Wed, Mar 13, 2024 at 12:35 AM Ion Gaztañaga via Boost wrote:

...I volunteer to be the review manager if the community agrees.

Robert, Vinnie, are you ok with this?

Yeah we can do that. Thanks

I think there is a more likely alternative:  Conditional acceptance. This is common recommendation of specific reviewers as well as the review manager. So it a large number of cases, the final result is is a library with changes made in accordance with information which comes about as part of the review process. I know this as having suffered through three reviews, (2 successful, 1 not) this process is a huge contributor to the quality of the final library. I would expect that this same result would obtain when applied to a new boost web site (as well as any boost tool). There is wide consensus that: a) Boost will benefit from an updated website. b) it's a big job that only Vinnie has been willing to undertake. So I believe that the effort has been and will be viewed positively. What I want is to see is wider participation if the submission and validation of ideas on functionaly, scope, and even aesthetics. I know this is a huge pain for the developers, just as it is for library developers. But the pain is worth it. It will result in a much better product. In the future when some raises the question: "Why don't we ..." the answer will be in the archived reviews thus short circuiting the endless circular discussion about these topics. Traditionally, boost tools (including the website) have been developed on some individual initiative by a member of the "inner circle" and presented as a "fait compli". I think we need a more open and formal process for these things. It's the Boost Way and it distinguishes us from the standards process. Robert Ramey On 3/12/24 7:42 PM, Vinnie Falco wrote:

On Tue, Mar 12, 2024 at 2:28 PM Robert Ramey via Boost wrote:

...

The authors of this work deserve a yes or no answer as to whether it's going to be accepted or not. That is the purpose of the Boost Review.

I think there is a more likely alternative:  Conditional acceptance. This is common

Thinking about this a bit more, I have some clarity into some of my recent uncomfortable feelings. A C++ library goes through the review process, and whether it passes or fails the library is still usable. The author can derive value from the library even if rejected, by using it in their own projects and encouraging others to use it and form a following.

A new Boost website however, is different. If the website is rejected, then most if not all of the work is for nothing. It is a total loss or close to it. We didn't build a new website on a whim, there was a discussion a few years ago about the need for a new site. There are stakeholders from the Boost community that have worked with us. Although the website was not designed by committee (if for no other reason that such a process makes failure likely), we did get community engagement once there was something worth looking at.

Unlike a C++ library, a new Boost website is really only usable for becoming a Boost website. The problem with a straight Boost Formal Review applied to the website, is that a rejection would likely discourage anyone from making the effort again in the first place. Case in point, if I was told at the beginning that our only option would be to publish our website on a different domain while David either co-opts our work himself to place on boost.org, or devalues our labor and puts up his own new website on boost.org, we would have never done the work in the first place.

The new website reflects my love the Boost Libraries, and in my opinion there needs to be an implicit understanding that if someone takes on the enormous risk of building a new Boost website, delivers on that promise, and demonstrates their understanding and adoption of Boost culture, we want to encourage and celebrate this. As the website (unlike a library) cannot be repurposed, a rejection would not only be an immense personal loss for me but represent a significant loss for Boost at a time when we need all the victories we can achieve.

I would ask everyone who loves Boost to rally and get behind this effort to publish this modern website, to show the onlookers and new potential community members that we are capable of supporting this high-risk / high-reward contribution. The alternative is too dismal to contemplate.

Thanks

-- Robert Ramey www.rrsd.com (805)569-3793