Hi Thomas, I've been tracking this conversation from afar but have not had much chance to get involved. Thank you for looking into this in depth. I will look into the trunk vs. release regression problems this weekend. It may take a little longer for me to look into the Iterator situation. Ron On May 26, 2010, at 12:58 PM, Thomas Klimpel wrote:
Hi,
The issues reporteted in <http://lists.boost.org/Archives/boost/2010/05/165798.php
are still open. Even so I wrote:
But from my point of view, the only responsibility I may have with respect to these issues is to notify the authors of the affected libraries, which I have done now.
, claiming that the bug is in Boost.MultiArray. However, my
. But there is definitively a "quasi unrelated" bug in Boost.MultiArray on trunk (and Thomas Witt might refer to this bug), as the "resize" regression test fails on nearly all compilers on
it seems that I wasn't yet successful in my attemps to notify the authors. The only one that has answered so far was Thomas Witt <http://lists.boost.org/Archives/boost/2010/05/165837.php personal belief is that the bug is in <http://svn.boost.org/svn/boost/trunk/boost/iterator/detail/facade_iterator_c... trunk but passes on release <http://lists.boost.org/Archives/boost/2010/05/165810.php
(the link contains a patch that fixes the failures).
I would be most happy, if
1) Ron (or somebody else that is allowed to apply patches to Boost.MultiArray) would simply fix the failure of the "resize" regression test on trunk (either by applying the provided patch or by fixing it in some other way).
2) Dave (or somebody else that has deep knowledge about Boost.Iterator) would confirm that the failures of Boost.MultiArray on MSVC-10 are caused by an issue in Boost.Iterator (and are probably related to Jeremy Siek's patch <https://svn.boost.org/trac/boost/changeset/21683
).
I guess the result if somebody will answer at all is that I will be asked to enter a track ticket for Boost.Iterator. So if this mail should stay without useful feedback, I will enter two track tickets for Boost.MultiArray and one track ticket for Boost.Iterator, in anticipation that the feedback would probably have suggested that I should do that.
Regards, Thomas