Go ahead.

On Mar 17, 2021, at 4:07 AM, Hans Dembinski via Boost wrote:

...

On 3/16/21 11:24 PM, Marshall Clow via Boost wrote: The master branch is is now open for post-beta merges, but only as described in the Post-Beta Merge Policy. See https://github.com/boostorg/boost/wiki/Releases%3A-Beta-Merge-Policy Reminder: The master branch closes for the release on Wednesday, Apr 7. Full calendar here: https://www.boost.org/development/ https://www.boost.org/development/

I would like to merge critical fixes implemented in Boost.Histogram develop to master:

https://github.com/boostorg/histogram/compare/master...develop

The change:

`accumulators::mean` and `accumulators::weighted_mean` now compute the correct variance when `operator+=` is called, for example, when histograms with accumulators are added; this was wrong before leading to too small variances

Unit tests were rewritten to catch this error and do have a more clear structure. Comments were added to code and some doxygen strings fixed. CI tests have passed.

Best regards, Hans _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

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

On Wed, Mar 17, 2021 at 2:10 PM Vinnie Falco wrote:

...

I'd like to merge Boost.JSON develop to master which contains Necessary Fixes from Very Reliable Sources

So yeah, about that merge... Thanks

On 25/03/2021 15:45, Marshall Clow wrote:

On Mar 25, 2021, at 8:06 AM, Niall Douglas via Boost wrote:

...

On 16/03/2021 20:24, Marshall Clow via Boost wrote:

...

The master branch is is now open for post-beta merges, but only as described in the Post-Beta Merge Policy. See https://github.com/boostorg/boost/wiki/Releases%3A-Beta-Merge-Policy

I would like to fix https://github.com/ned14/outcome/issues/249 by improving the logic at https://github.com/ned14/outcome/blob/develop/include/outcome/convert.hpp#L3... please.

Do you have a commit that fixes this?

The change is trivial. Current: ```c++ #if !defined(_MSC_VER) && !defined(__clang__) && (__GNUC__ < 9 || __cpp_concepts < 201907L) #define OUTCOME_GCC6_CONCEPT_BOOL bool #else #define OUTCOME_GCC6_CONCEPT_BOOL #endif ``` To be replaced with: ```c++ #if (defined(_MSC_VER) || defined(__clang__) || (defined(__GNUC__) && __cpp_concepts >= 201707) || OUTCOME_FORCE_STD_CXX_CONCEPTS) && !OUTCOME_FORCE_LEGACY_GCC_CXX_CONCEPTS #define OUTCOME_GCC6_CONCEPT_BOOL #else #define OUTCOME_GCC6_CONCEPT_BOOL bool #endif ``` To explain, GCC expects bool with concept, or not, in varying configuration combinations which have evolved over GCC versions, including apparently point releases, which is deeply unhelpful. This has led to repeated bug reports for not just Outcome, but also for ASIO and many other C++ projects. If this above fix doesn't permanently shut this constant source of bug reports, I'll be permanently disabling legacy GCC concepts support in Outcome. I couldn't be arsed with supporting how unpredictably broken GCC is with this. Niall

On 25/03/2021 15:45, Marshall Clow wrote:

On Mar 25, 2021, at 8:06 AM, Niall Douglas via Boost wrote:

...

On 16/03/2021 20:24, Marshall Clow via Boost wrote:

...

The master branch is is now open for post-beta merges, but only as described in the Post-Beta Merge Policy. See https://github.com/boostorg/boost/wiki/Releases%3A-Beta-Merge-Policy

I would like to fix https://github.com/ned14/outcome/issues/249 by improving the logic at https://github.com/ned14/outcome/blob/develop/include/outcome/convert.hpp#L3... please.

Do you have a commit that fixes this?

Judging from https://www.boost.org/development/tests/develop/developer/outcome.html, the commit fixing this shown at https://github.com/boostorg/outcome/commits/develop has not broken anything. May I merge this to master? Niall

On 01/04/2021 16:39, John Maddock wrote:

On 16/03/2021 20:24, Marshall Clow via Boost wrote:

...

The master branch is is now open for post-beta merges, but only as described in the Post-Beta Merge Policy. See https://github.com/boostorg/boost/wiki/Releases%3A-Beta-Merge-Policy

I'd like to merge this change: https://github.com/boostorg/math/pull/593 to master if possible.

Explanation: parallel execution of some algorithms has been newly added to this release, but we've found one (Gini coefficient calculation) to be faulty, this PR is a temporary sticking plaster that directs everything to the serial algorithm.

The one CI failure is spurious (I haven't yet merged a CircleCI config to develop, much less master, but it is enabled for the repository while I work on getting that ready for develop).

Many thanks, John.

Ping? -- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

On Fri, Apr 2, 2021 at 6:40 AM Andrey Semashev via Boost wrote:

I'd like to merge the following PR to Boost.Atomic master:

https://github.com/boostorg/atomic/pull/49

This fixes compilation with MSVC for ARM.

CI tests have passed.

Go ahead and merge. Glen

I would like to merge the following PR to Boost.Log master:

https://github.com/boostorg/log/pull/149

It fixes incompatibility with MinGW-w64 when building against UCRT.

And I'd like to merge basically the same fix for Nowide: https://github.com/boostorg/nowide/pull/130

On Tue, 6 Apr 2021 at 15:53, Richard Hodges wrote:

On Tue, 16 Mar 2021 at 21:25, Marshall Clow via Boost < boost@lists.boost.org> wrote:

...

The master branch is is now open for post-beta merges, but only as described in the Post-Beta Merge Policy. See <https://github.com/boostorg/boost/wiki/Releases%3A-Beta-Merge-Policy

...

Reminder: The master branch closes for the release on Wednesday, Apr 7. Full calendar here: https://www.boost.org/development/ < https://www.boost.org/development/>

I'll need to make one commit later on today. This will be limited to correcting an incorrect assumption in one test and only affects the reliability of the green build indicator.

I think it's important to squeeze this one in as otherwise we might undermine confidence in the library after the release.

Sorry for the late notice.

Once again I feel obliged to apologise to the release managers. I inadvertently rebased my changes onto master rather than develop before seeking approval for updating beast master. No production code is affected. Commit details are here: Fix a test to make it deterministic https://github.com/boostorg/beast/commit/7257d64a2179621d5514a310472ad28db49... Compensate for the upstream removal of older compilers in the ubuntu GHA image https://github.com/boostorg/beast/commit/fa1004fcbfaa9c8b654115438aa7efc9d76... Set version number https://github.com/boostorg/beast/commit/83aaefaf725d2dd125132c40a3f26811979...