May 2023 - Boost - lists.boost.org

Re: [boost] Q1 2023 Update: C++ Alliance's Efforts (PDF Attached)
by Julien Blanc 03 May '23

03 May '23

Le 2023-05-02 15:25, David Sankel via Boost a écrit : > I wanted to call out Sam Darwin's work on this: > >> Transitioned boost mailing list gateway from mailgun to Amazon SES > > This transition has resulted in a $575 per month cost reduction in > Boost > Foundation hosting expenses. Sam did an excellent job here. Did that happen around end of february / start of march? I've noticed that threaded view was broken around that date (and still is). At first, i blamed my client (evolution) but i see the same issue with another (roundcube). A quick look at mail headers seems to show wrong values in the References: header (but that is far beyond my expertise). > Thanks Sam and C++ Alliance for this valuable work! Yes, thanks for all this invisible work. Julien

1 0

Re: [boost] Replacing the ML with a forum
by Dominique Devienne 02 May '23

02 May '23

On Mon, May 1, 2023 at 8:15 AM Vinnie Falco via Boost <boost(a)lists.boost.org> wrote: > On Sun, Apr 30, 2023 at 8:56 PM Rainer Deyke via Boost > <boost(a)lists.boost.org> wrote: > > The SDL mailing list moved from true mailing list on gmane to Discourse > > web forum with email notification (but no gmane) a while back > > Good thing that's not happening to Boost > I've already mentioned it, but I feel compelled to re-iterate, given the lets-built-it direction, although I know it's most likely in vein. SQLite made the transition from ML to a custom-built Forum, and it went/was OK. I and several others were against it. In the end, the community survived and thrived. I've used Discord in the past. I've used Basecamp. I've used News (Google Reader, RIP). And of course I've used MLs too, these ones included. [The Fossil Forums][1] are low-key, fast, i.e. good enough. And very low resources on the backend. The Forums integrate with [Pikchr][2] for text-based diagramming. Email notification is implemented, but replies by emails are not. You don't need to use Fossil SCM to use them. Many people use Fossil just for the Forums. Unlike SQLite, Fossil is open to contributions too, although the community is opinionated. I have no stake in Fossil. I'm just a happy user. FWIW. --DD [1]: https://www.fossil-scm.org/home/doc/trunk/www/forum.wiki [2]: https://fossil-scm.org/home/doc/trunk/www/pikchr.md

2 1

Re: [boost] The History of Boost
by Vinnie Falco 02 May '23

02 May '23

On Mon, May 1, 2023 at 1:44 PM Andrey Semashev via Boost <boost(a)lists.boost.org> wrote: > I think, you can start by copying all the existing information on the > current site and then improve on it. I don't see the point in > registering issues requesting adding information that is already there > on the current website. Yep! Well, maybe just one issue with links to all the existing info? > IMO, this style of navigation is too obscure. I would have never thought > to click on "User Guide" in the bottom corner of the screen to switch to > other documents. That UI is temporary, it is what comes with the "Antora Default UI." In the actual website there will be a nice looking page showing a handful of cards, one for each document, with information about what is inside the document and a small sample of the most important topics inside. And you will always be able to get to this page just by clicking "Learn" in the top level navigation. > The navigation line at the top is useless because > "home" leads to "User Guide" and not a landing page with contents. Yes, that top level navigation area is temporary. In the actual website one of the top level navigation items says "Learn" and this takes you to a landing page that has beautiful cards which link to each individual document. No matter where you are on the website (which includes individual Boost library documentation) it will be possible to reach the landing page with the site-docs in a single click (by pressing "Learn" at the top). The restyling of the Antora UI to match the website is not a trivial undertaking; all of the Asciidoc CSS classes have to be styled with it, and there are a LOT of them. Lists, tables, admonitions, source code boxes (especially those because we want them to be awesome the whole focus of our website is on source code). This is happening in parallel with the authoring of the exposition. Thanks!

1 0

Re: [boost] Boost CI - a modest proposal
by Hans Dembinski 02 May '23

02 May '23

> On 30. Apr 2023, at 17:37, Robert Ramey via Boost <boost(a)lists.boost.org> wrote: > > the implementation of Boost CI follows the normal pattern for the implementation of this functionality. This entails download the latest version of the library to be tested and it's dependencies (or maybe all of boost?) from the repo, Building everything from scratch, running all the tests and logging the output there of. This is done for each compiler / environment to be tested. Other than the raw console logs, there is no presentation of the tests results (e.g boost test matrix). > > This is extremely unsatisfactory. > > a) It's ridiculously resource intensive. Rebuilding everything all the time. > > b) takes forever unless one has many more servers than we have. These two points can be addressed somewhat by cutting down the matrix and by using ccache. You don't really need to test all versions of all compilers. For daily work it is sufficient to use the lowest and highest compiler versions you want to support. If the main time is spend in compiling the test instead of running them, Boost.CI could be sped up by using ccache. If have used ccache in other projects, where it sped up the CI considerably. Best regards, Hans

1 0

Re: [boost] Boost.Compat
by Hans Dembinski 02 May '23

02 May '23

> On 30. Apr 2023, at 01:06, Andrey Semashev via Boost <boost(a)lists.boost.org> wrote: > > I really don't like us having > multiple copies of the same components for the sole purpose of reducing > dependencies (in the case of latch - reducing tests dependencies, not > even library dependencies). Me neither, but like everything in life it is a trade-off. I want to both reduce duplication and dependencies in Boost.Histogram, which is a rather high-level library that builds on a lot of low-level Boost functionality. To reduce dependencies, I used to have my own implementation of stuff like variant or span, which I eventually replaced with boost::span in Boost.Core and Boost.Variant2. This is partially repeating what others have said: I see utility in Boost.Compat, because the scope of Boost.Compat is well-defined, while the scope of Boost.Core isn't. I have had discussions with the Boost.Core devs over what should be in Boost.Core or not. Knowing the scope of a boost library helps both with making requests that have a chance to be accepted and also makes it more clear for users what to expect from this library. Regarding the reviewing of Boost.Compat: A light review is sufficient, since we can skip the discussion of the design/scope/etc, but it looks like we want to have some initial review. We are currently discussing scope and rules for this library, these should be approved by a formal review. Just my two cents, Hans

1 0

Re: [boost] Boost CI Infrastructure
by Niall Douglas 01 May '23

01 May '23

On 30/04/2023 18:31, Vinnie Falco via Boost wrote: > Alan de Freitas has done a considerable amount of good work in this > area so if you would like to ask specific questions on the list or in > the Slack workspace that is also a good choice (or you can ask me or > Sam). Alan contributed a whole bunch of GHA CI for Boost.Outcome. It runs very well, it usually completes before the Drone CI does for Outcome. People here may be interested in studying it and how it works. My thanks to Alan for doing all the heavy lifting there. And to Peter Dimov for identifying things which had become broken in Boost.Outcome due to changes elsewhere in Boost but hadn't registered on CI, and therefore needed much wider CI testing (which Alan then implemented). Also, thanks to Vinnie for asking Alan to do that work for Outcome. I haven't had the time recently (spent the last three days driving around a six tonne digger lifting 'lego' concrete blocks, it has been very tedious). Niall

1 0

Re: [boost] Boost.Compat
by Christian Mazakas 01 May '23

01 May '23

> It doesn't really matter how many dependencies Boost.Thread adds. What > matters is how expensive <boost/thread/latch.hpp> is. Personally, I > would have no problem depending on it, if I really needed this component. One problem here is that Thread is effectively unmaintained. In general, I'm not a huge fan of depending on libraries like this in my own tests because Unordered has a very strict set of flags and many Boost libraries won't clear the same bar. I remember in the past having a test dependency on one of the bigger Boost libraries and it wasn't exactly painless. I can't remember now if it was Container or Filesystem. Boost.Compat has utility in that it's essentially just a collection of substitutes for all the goodies the committee added that could've been in C++11. This benefits both authors and users massively, instead of purporting that anyone should rely on an unmaintained lib. - Christian

1 0

Re: [boost] The History of Boost
by Andrey Semashev 01 May '23

01 May '23

On 5/1/23 18:36, Vinnie Falco via Boost wrote: > Now I'm not asking anyone to actually write docs, but I notice there > has been basically no participation from anyone on the mailing in the > site-docs that I announced recently, with the exception of a couple of > very helpful people who also happen to be staff engineers at C++ > Alliance. I think, you can start by copying all the existing information on the current site and then improve on it. I don't see the point in registering issues requesting adding information that is already there on the current website. > Reminder: > > Site Documentation (there are 4 docs here, use the control at the > bottom left to switch from User's Guide to the others): > > https://docs.cppalliance.org/doc/user-guide IMO, this style of navigation is too obscure. I would have never thought to click on "User Guide" in the bottom corner of the screen to switch to other documents. The navigation line at the top is useless because "home" leads to "User Guide" and not a landing page with contents.

1 0

Q1 2023 Update: C++ Alliance's Efforts (PDF Attached)
by Louis Tatta 01 May '23

01 May '23

Dear Boost Libraries Mailing List, I am writing to share with you a summary of the efforts of our team at C++ Alliance to enhance the Boost Libraries during Q1 2023. Please find attached a PDF document that provides an overview of the work we have done to maintain existing libraries, develop new ones, and improve documentation and support. We are proud of the progress we have made so far and remain committed to improving the Boost Libraries in the future. We believe that these libraries can help C++ developers write more efficient, reliable, and maintainable code. Thank you for your time and attention, and please let us know if you have any feedback or questions. Best regards, Louis C++ Alliance Team

1 0

The History of Boost
by Vinnie Falco 01 May '23

01 May '23

Now I'm not asking anyone to actually write docs, but I notice there has been basically no participation from anyone on the mailing in the site-docs that I announced recently, with the exception of a couple of very helpful people who also happen to be staff engineers at C++ Alliance. Reminder: Site Documentation (there are 4 docs here, use the control at the bottom left to switch from User's Guide to the others): https://docs.cppalliance.org/doc/user-guide GitHub issues for site-docs: https://github.com/cppalliance/site-docs/issues You don't need to submit a pull request; adding information to existing issues, or opening new issues is an enormous help. Now on that subject we want to make sure we have a "History Of Boost" chapter which thoughtfully outlines the founding of Boost and its noteworthy historical events leading to the present day. We want to make sure we get this right and credit everyone who needs crediting, and not miss anything important. Boost is volunteer work and since there is no financial compensation the only other currency is social credit and reputation so this has to be correct. If you could please visit this issue and add any information that you feel is relevant to the history of Boost (it can be notes or shorthand don't worry about grammar, Peter Turcan will edit it down and add it) that would be of enormous help: Issue on The History of Boost: https://github.com/cppalliance/site-docs/issues/37 Of course adding to any other issue or opening new ones is a help too, so feel free! Thanks

1 0