On Fri, Nov 20, 2015 at 12:44 PM, Peter Uhn��k <i.uhnak@gmail.com> wrote:

And not just announcers, some of the tests there leave quite a footprint.

Indeed some tests are not designed to be silent.

I've recently discussed something similar with regards to performance
of subclassing in tests
http://forum.world.st/fast-subclassing-class-creation-in-tests-td4857779.html
So Kernel-Tests might also benefit from it.

Right now I've also encountered this thread
http://forum.world.st/New-class-factory-for-tests-and-other-use-cases-td1297663.html#a1297667
, it's quite old (2009), but maybe it could be used instead of
directly manipulating the env.

On the other hand, how often do you manually trigger Kernel-Tests? CI
doesn't care that much about it.

Our current issue is that we need to update something in an application based on system announcers.

The fact that tests trigger events generates extra work which makes some builds longer.

Changing the tests would take too much time, so we'll look for alternatives.

Cheers,

Andrei��

��

Peter

On Fri, Nov 20, 2015 at 12:21 PM, Andrei Chis
<chisvasileandrei@gmail.com> wrote:
> Hi,
>
> Ideally shouldn't Kernel-Tests, or tests in general, be silent and not
> trigger a lot of system announcers?
> Right now Kernel-Tests trigger around 250 announcements through the
> SystemAnnouncer.
>
> Cheers,
> Andrei