Hi,
kytv wrote (02 Aug 2015 20:12:11 GMT) :
> I started a run with the test suite using
> tails-i386-bugfix_9748-revert-apt-amd64-1.5-20150801T1933Z-ac29b77+devel@???
> but tests which use bridges fail with something like
> [...]
> On the screen is "Tor Network Settings" window. Basically the
> "Configure" button wasn't pressed.
Can this be reproduced, or was it a one-time error?
Can you reproduce with an ISO built from devel on Jenkins?
> These same tests with my ISO with bugfix/9748-revert-apt-amd64 merged
> into devel will pass. The log looks like the following:
> f5c0695 Merge remote-tracking branch 'origin/bugfix/9748-revert-apt-amd64' into devel
> ac29b77 Revert inclusion of syslinux:amd64 in the ISO.
> 8877514 Merge branch 'stable' into devel
> Note that the commits (other than the merge commit) match what jenkins
> used for its ISO (ac29b77+devel@8877514)
So in the end we should indeed have the same code in the Git tree used
to build the ISO.
> My experience seems to differ from the intention outlined at
> https://tails.boum.org/blueprint/automated_builds_and_tests/autobuild_specs/#index2h2
>> A topic branch may be lagging behind the base branch it's based upon.
>> What we're interested in is generally not whether a (possibly
>> outdated) topic branch builds fine, but whether it would build fine
>> once merged into its base branch
> In this case the ISO was indeed successfully built by jenkins but
> the end result was very different from what I got I merged the branches
> locally and built an ISO.
In Jenkins we merge the base branch into the topic one, not the
opposite. No idea if it changes anything in this specific case, in
general it should not.
I suspect that what you're seeing has little to do with what's in Git,
and more to do with what's in Debian, or gzip vs. xz compression, or
something else.
Cheers,
--
intrigeri
