On 10/16/15 9:25 AM, Tom Kent wrote:
On Fri, Oct 16, 2015 at 10:15 AM, Robert Ramey <ramey@rrsd.com> wrote:
Would it be possible to revert the changes to develop until you have a better handle on them, so that the runners can continue for all the other projects?
I'm not sure this would help. Except for one minor change in the examples which shouldn't affect the testing, I believe that develop and master are in sync. Since windows tests on the master are quite old it may be the problem exists on the master but just aren't showing up there (yet). I'm reluctant to starting doing any git jujitzu at this point as I'm trying to get everything hunky dory before next wednesday when a 1.60 cut off is planned.
This would also rule-out the problem actually being a change in some other library, that is just showing up here.
I don't know if this is easy/possible with the Boost Testing setup but when I run here, I can just switch and particular library between branches and run the tests. That is, you might be able to locally just switch the serialization library to master and run the tests. Since this likely will give the same results, you'd actually have to switch the master to the 1.5 release version. Rest assured that, as we speak, I'm trying to get out the last nits on the next version of the serialization library so hopefully the whole question will be moot soon. I also believe that there is a pending issue related to visibility on GCC plaforms but I have to get other stuff fixed before I can look into this one. Thanks for spending time on this. Your work is very helpful and critically important to the success of Boost. Robert Ramey
Tom
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost