I have reviewed this request. tl/dr: I recommend approval.
I have looked at the debdiff and the changes seem to be as described in the bug report. Almost all of the things described sound very much like important bugfixes. I'm not entirely sure about "Improvements in key parts of the consistency engine" but the approach taken by upstream seem reassuring and if I were the maintainer I think I would choose to take that change too rather than trying to drop it. I did notice that the debdiff contains a fair few test suite changes, including notably what looks like new tests. I am comfortable with that, even though this was not called out in the bug or the changelog. I guess that's just upstream practice. The risk of accepting is course is that there is some bug in these changes - perhaps even a data loss bug. But the fact that this code has been in the upstream stable branch for a month mitigates against that, and the known problems described seem worse than the cure. I did not do a code review in context, and I'm not familiar with the codebase, so I'm not sure that the changes are indeed right. Such a detailed code review exercise seems beyond our realistic capabilities for a stable update. I'm not an ftpmaster and have no formal status. I'm just hoping to be helpful. Ian. -- Ian Jackson <ijack...@chiark.greenend.org.uk> These opinions are my own. If I emailed you from an address @fyvzl.net or @evade.org.uk, that is a private address which bypasses my fierce spamfilter.