There is another pull request

https://github.com/boostorg/filesystem/pull/43

Can this maybe be taken into account for boost 1.66 ?

Thanks !

On 25 August 2017 at 11:05, Olivier Tristan via Boost-users <boost-users@lists.boost.org> wrote:

Hi Guys,

I would like to bump this particular issue that is quite annoying for me.

Basically it crashes as soon as you call boost::filesystem::exists in the ctor of a static variable initialized at program startup

https://svn.boost.org/trac10/ticket/12987

Looks like there was a fix 4 month ago but this hasn't been pushed into the trunk yet

Any idea how to push the process forward ?

Thanks !

-- 
Olivier Tristan
Research & Development
www.uvi.net

I also hit this and fixed it in, I believe, a simpler way:

https://github.com/boostorg/filesystem/pull/50

(Also note that as long as the object is static and in a namespace,

static initialization fiasco will happen anyway, which is why I simply moved it in the scope of the function.)

But as you can see there is no response there either yet.

I suspect that work on making boost.filesystem match std::filesystem is

taking all the author's available resources.

Joël Lamotte

 

_______________________________________________
Boost-users mailing list
Boost-users@lists.boost.org
https://lists.boost.org/mailman/listinfo.cgi/boost-users

-- 
Olivier Tristan
Research & Development
www.uvi.net