Hey,
We have quite a few Debian Continuous Integration [1] test failures.
Some of those failures were caused by minor testing issues [2] that
don't affect package usability and can be easily fixed. However, I
wonder what is our current policy about those failures:
1. Should we file a bug report for each failed package (regardless of
severity)?
2. Should we treat those non-critical test failures as bugs that can ask
for an unblock request or should we just treat them as minor issues that
can be fixed after freeze?
For example, there is a CI failure in my ruby-beautify package. Since a
new upstream version was recently released, I was just going to fix that
issue when packing a new version (targeted @ experimental). Is that ok
or should I report a bug, fix it in current version (and request
unblock) and only then pack a new version?
What about all the other
packages that don't have a new upstream? Should
we fix whenever possible and upload to experimental or just wait and
look into it after freeze ends?
Regards,
T.
[1]
https://qa.debian.org/developer.php?login=pkg-ruby-extras-maintainers@lists.alioth.debian.org
[2] Most of the bugs reported during autopkgtest fall into few
categories, some of which are non-critical (or simply testing
environment issues). A full report about those will be posted when
ready. At the moment you can check it out at gobby.d.o.