[Development] failing network tests on CI

[Knoll Lars]

On 8/8/13 6:16 PM, "Laszlo Papp" <lpapp at kde.org> wrote:

>On Thu, Aug 8, 2013 at 5:08 PM, Thiago Macieira
><thiago.macieira at intel.com> wrote:
>
>On quinta-feira, 8 de agosto de 2013 17:49:11, Stephen Kelly wrote:
>> > If we disable the tests, to be reasonably sure of not introducing
>> > regressions, we should also stop all changes to QtNetwork, event loops
>> > (QtCore and QPA plugins) and QObject kernel.
>>
>> That's not necessary. The QtMetrics site can be checked regularly or
>>before
>> release if needed.
>
>
>Past experience shows that this does not work. It won't be checked, and
>even
>if it is checked, all the issues won't be fixed in time.
>
>If we disable the tests, let's do it with open eyes: there will be
>regressions
>and we'd be accepting them.
>
>
>
>I agree with Thiago here.

Yes, he's right. We can do manual work trying to avoid any regressions,
but they are likely to happen. Still if we can't get the server fixed,
it's IMO better to mark these tests as insignificant then to have many
failed CI runs.

Cheers,
Lars