[boost-git] fiber not contained?
Hi, I was asked several times why boost.fiber is not contained if boost's git-repo is cloned with git clone --recursive http://github.com/boostorg/boost.git. I'm wondering because boost.fiber is already included in the regression tests (master/develop). Oliver
Hi Oliver, Am Montag, 27. Juni 2016, 09:41:05 CEST schrieb Oliver Kowalke:
I was asked several times why boost.fiber is not contained if boost's git-repo is cloned with git clone --recursive http://github.com/boostorg/boost.git.
This will checkout master and Boost.Fiber is not in the submodule list of the master branch. It is also missing in the last release tarball.
I'm wondering because boost.fiber is already included in the regression tests (master/develop).
This is strange. I can confirm that it is seems to be in both matrices. Something broken in the regression testing? Yours, Jürgen -- * Dipl.-Math. Jürgen Hunold ! * voice: ++49 4257 300 ! Fährstraße 1 * fax : ++49 4257 300 ! 31609 Balge/Sebbenhausen * jhunold@gmx.eu ! Germany
On Mon, Jun 27, 2016 at 4:56 AM, Jürgen Hunold <jhunold@gmx.eu> wrote:
Am Montag, 27. Juni 2016, 09:41:05 CEST schrieb Oliver Kowalke:
I was asked several times why boost.fiber is not contained if boost's git-repo is cloned with git clone --recursive http://github.com/boostorg/boost.git.
This will checkout master and Boost.Fiber is not in the submodule list of the master branch. It is also missing in the last release tarball.
Boost.Fiber was accepted more recently than the release of Boost 1.61. I believe Oliver is asking: what must happen for Boost.Fiber to be added to the submodule list of the master branch of the superproject?
On Mon, Jun 27, 2016 at 2:41 AM, Oliver Kowalke <oliver.kowalke@gmail.com> wrote:
I was asked several times why boost.fiber is not contained if boost's git-repo is cloned with git clone --recursive http://github.com/boostorg/boost.git.
I'm wondering because boost.fiber is already included in the regression tests (master/develop).
Hm.. You are correct in that fiber is not in the master branch yet. Usually one of the release managers adds it before a release. But it's definitely strange that it's getting tests posted in the master branch results. And I don't know why it would do that. I'll need to investigate. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail
On Mon, Jun 27, 2016 at 12:32 PM, Rene Rivera <grafikrobot@gmail.com> wrote:
On Mon, Jun 27, 2016 at 2:41 AM, Oliver Kowalke <oliver.kowalke@gmail.com> wrote:
I was asked several times why boost.fiber is not contained if boost's git-repo is cloned with git clone --recursive http://github.com/boostorg/boost.git.
I'm wondering because boost.fiber is already included in the regression tests (master/develop).
Hm.. You are correct in that fiber is not in the master branch yet. Usually one of the release managers adds it before a release. But it's definitely strange that it's getting tests posted in the master branch results. And I don't know why it would do that. I'll need to investigate.
PS. And now I know why it's happening. Seems the testers are running develop & master tests from the same testing tree. And hence the same git clone. Which doesn't work the way regression script does the clone & checkout. As it leaves submodules around when it checks out from one branch to another. As git refuses to delete empty submodule directories. So I'll have to find a way to deal with this use case :-( PPS. I hate git submodules. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail
2016-06-27 20:46 GMT+02:00 Rene Rivera <grafikrobot@gmail.com>:
PS. And now I know why it's happening. Seems the testers are running develop & master tests from the same testing tree. And hence the same git clone.
OK
Which doesn't work the way regression script does the clone & checkout. As it leaves submodules around when it checks out from one branch to another. As git refuses to delete empty submodule directories. So I'll have to find a way to deal with this use case :-(
Could you add boost.fiber to master branch/1.62 release, please?
On Tue, Jun 28, 2016 at 12:47 AM, Oliver Kowalke <oliver.kowalke@gmail.com> wrote:
2016-06-27 20:46 GMT+02:00 Rene Rivera <grafikrobot@gmail.com>:
Which doesn't work the way regression script does the clone & checkout. As it leaves submodules around when it checks out from one branch to another. As git refuses to delete empty submodule directories. So I'll have to find a way to deal with this use case :-(
Could you add boost.fiber to master branch/1.62 release, please?
Probably :-) (I'm always fuzzy about the git commands for this). But I still have to fix the testing issue, as fiber is not the only library affected by this. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail
On Tue, Jun 28, 2016 at 12:47 AM, Oliver Kowalke <oliver.kowalke@gmail.com> wrote:
2016-06-27 20:46 GMT+02:00 Rene Rivera <grafikrobot@gmail.com>:
PS. And now I know why it's happening. Seems the testers are running develop & master tests from the same testing tree. And hence the same git clone.
OK
FYI; This should now be fixed. This is yet another example of git not truly dealing with submodules appropriately. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail
2016-06-28 23:17 GMT+02:00 Rene Rivera <grafikrobot@gmail.com>:
FYI; This should now be fixed. This is yet another example of git not truly dealing with submodules appropriately.
OK, thank you. May I remind you to add library fiber to branch master? I'd like to get the master regression tests too. best regards, Oliver
participants (4)
-
Jürgen Hunold -
Nat Goodspeed -
Oliver Kowalke -
Rene Rivera