On 8/9/22 01:27, Marshall Clow via Boost wrote:

On Aug 8, 2022, at 7:25 AM, Andrey Semashev via Boost wrote:

...

On 8/4/22 16:25, Marshall Clow via Boost wrote:

...

The first release candidates for the 1.80.0 release are now available at: https://boostorg.jfrog.io/artifactory/main/release/1.80.0/source/

The SHA256 checksums are as follows:

4b2136f98bdd1f5857f1c3dea9ac2018effe65286cf251534b6ae20cc45e1847 boost_1_80_0_rc1.tar.gz 1e19565d82e43bc59209a168f5ac899d3ba471d55c7610c677d4ccf2c9c500c0 boost_1_80_0_rc1.tar.bz2 d12a2af721e22dbfb984ef061ee4c4ab2387f1904f6d65bc5acebaa34d6366ec boost_1_80_0_rc1.7z e34756f63abe8ac34b35352743f17d061fcc825969a2dd8458264edb38781782 boost_1_80_0_rc1.zip

As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy.

I would like to merge this last minute fix to Boost.Filesystem master:

https://github.com/boostorg/filesystem/commit/9c9d127bddc2b72187c57f4933c496...

This should fix error constructing a directory iterator over a network share on Windows prior to 10.

Is this a regression against 1.79.0?

The problem was introduced in 1.79.0, so a regression since 1.78.0.

On Aug 9, 2022, at 10:17 AM, Andrey Semashev via Boost wrote:

On 8/9/22 02:48, Andrey Semashev wrote:

...

On 8/9/22 01:27, Marshall Clow via Boost wrote:

...

On Aug 8, 2022, at 7:25 AM, Andrey Semashev via Boost wrote:

...

On 8/4/22 16:25, Marshall Clow via Boost wrote:

...

The first release candidates for the 1.80.0 release are now available at: https://boostorg.jfrog.io/artifactory/main/release/1.80.0/source/

The SHA256 checksums are as follows:

4b2136f98bdd1f5857f1c3dea9ac2018effe65286cf251534b6ae20cc45e1847 boost_1_80_0_rc1.tar.gz 1e19565d82e43bc59209a168f5ac899d3ba471d55c7610c677d4ccf2c9c500c0 boost_1_80_0_rc1.tar.bz2 d12a2af721e22dbfb984ef061ee4c4ab2387f1904f6d65bc5acebaa34d6366ec boost_1_80_0_rc1.7z e34756f63abe8ac34b35352743f17d061fcc825969a2dd8458264edb38781782 boost_1_80_0_rc1.zip

As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy.

I would like to merge this last minute fix to Boost.Filesystem master:

https://github.com/boostorg/filesystem/commit/9c9d127bddc2b72187c57f4933c496...

This should fix error constructing a directory iterator over a network share on Windows prior to 10.

Is this a regression against 1.79.0?

The problem was introduced in 1.79.0, so a regression since 1.78.0.

Gentle ping. 1.80 release is closing in, so if we're going to allow this change in, please let me know ASAP.

I talked to the other release managers, and we’re thinking .. not. The reasons are: * This has been this way since April, and no one noticed until last week. * Your checkin comment does not exactly fill me with confidence: “probably fixes #246” * Windows 10 was released in 2015. This just doesn’t seem like a sufficient reason to roll an RC2. — Marshall

On Aug 9, 2022, at 10:38 AM, Andrey Semashev via Boost wrote:

On 8/9/22 20:30, Marshall Clow wrote:

...

On Aug 9, 2022, at 10:17 AM, Andrey Semashev via Boost mailto:boost@lists.boost.org> wrote:

...

On 8/9/22 02:48, Andrey Semashev wrote:

...

On 8/9/22 01:27, Marshall Clow via Boost wrote:

...

On Aug 8, 2022, at 7:25 AM, Andrey Semashev via Boost mailto:boost@lists.boost.org> wrote:

...

On 8/4/22 16:25, Marshall Clow via Boost wrote: > The first release candidates for the 1.80.0 release are now > available at: > <https://boostorg.jfrog.io/artifactory/main/release/1.80.0/source/ > https://boostorg.jfrog.io/artifactory/main/release/1.80.0/source/> > > The SHA256 checksums are as follows: > > 4b2136f98bdd1f5857f1c3dea9ac2018effe65286cf251534b6ae20cc45e1847boost_1_80_0_rc1.tar.gz > 1e19565d82e43bc59209a168f5ac899d3ba471d55c7610c677d4ccf2c9c500c0boost_1_80_0_rc1.tar.bz2 > d12a2af721e22dbfb984ef061ee4c4ab2387f1904f6d65bc5acebaa34d6366ecboost_1_80_0_rc1.7z > e34756f63abe8ac34b35352743f17d061fcc825969a2dd8458264edb38781782boost_1_80_0_rc1.zip > > > As always, the release managers would appreciate it if you > download the > candidate of your choice and give building it a try. Please report > both > success and failure, and anything else that is noteworthy.

I would like to merge this last minute fix to Boost.Filesystem master:

https://github.com/boostorg/filesystem/commit/9c9d127bddc2b72187c57f4933c496... https://github.com/boostorg/filesystem/commit/9c9d127bddc2b72187c57f4933c496...

This should fix error constructing a directory iterator over a network share on Windows prior to 10.

Is this a regression against 1.79.0?

The problem was introduced in 1.79.0, so a regression since 1.78.0.

Gentle ping. 1.80 release is closing in, so if we're going to allow this change in, please let me know ASAP.

I talked to the other release managers, and we’re thinking .. not.

The reasons are: * This has been this way since April, and no one noticed until last week. * Your checkin comment does not exactly fill me with confidence: “probably fixes #246” * Windows 10 was released in 2015.

This just doesn’t seem like a sufficient reason to roll an RC2.

Ok.

Re. "Likely fixes" comment, I wrote that because I cannot test it myself and one of the original reporters haven't responded (yet). The other person who suggested the patch also had this problem and fixed it with the patch, so I'm reasonably confident in it.

Alas, it seems to be the tradition now for multiple problem reports to arrive during the release cycle.

It does. And we (the release managers) discuss each one, and decide if it (or “they”, in aggregate) are serious enough to roll another release candidate. This involves work for the release managers, but not *that* much. However, all the people who test the RCs then re-run their tests as well. — Marshall