Am 01/31/17 um 01:11 schrieb Stephen Connolly:
> So this sounds like the original tests are wrong.
Correct. They do not catch parent resolution failures although they
should have detected them.
>
> As such we shipped broken versions of Maven. It wasn't ever our intent that
> those projects work
So this sounds like the original tests are wrong.
As such we shipped broken versions of Maven. It wasn't ever our intent that
those projects work with 3.3.x rather a bug in the tests.
In that regard *my* personal pref is that all these tests have the range
starting from [3.2.2,) which would mean
Am 01/30/17 um 23:19 schrieb Stephen Connolly:
> I find it hard to follow your explanations as to what the change is, but I
> think I understand...
>
> could you try and rephrase what your change is because the current phrasing
> is rather more alarming than I suspect it is
1. The commit to the c
(I am trying to confirm my understanding of the diff:
https://github.com/apache/maven-integration-testing/compare/master...MNG-2199
On 30 January 2017 at 22:19, Stephen Connolly <
stephen.alan.conno...@gmail.com> wrote:
> I find it hard to follow your explanations as to what the change is, but I
I find it hard to follow your explanations as to what the change is, but I
think I understand...
could you try and rephrase what your change is because the current phrasing
is rather more alarming than I suspect it is
On 30 January 2017 at 18:34, Christian Schulte wrote:
> Hi,
>
> I'd like to m
Hi,
I'd like to merge the MNG-2199 branch to master for 3.5.0. I think this
should really go into 3.5.0 and not 3.5.1. Anyone second that? I tried
to explain things as much as possible in the commit messages. There also
is one IT which needs to be deprecated because it is testing an invalid
projec
