<br><br><div class="gmail_quote">On Wed, Sep 26, 2012 at 2:55 AM, Vijay Bellur <span dir="ltr"><<a href="mailto:vbellur@redhat.com" target="_blank">vbellur@redhat.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="HOEnZb"><div class="h5">On 09/26/2012 02:52 PM, Deepak C Shetty wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On 09/26/2012 11:41 AM, Vijay Bellur wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On 09/26/2012 10:34 AM, Deepak C Shetty wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On 09/25/2012 04:13 PM, Vijay Bellur wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Hi All,<br>
<br>
We intend to bring the following change in our gerrit based workflow:<br>
<br>
- Introduce +2 and -2 for Verified in Gerrit<br>
- +2 for Verified to be necessary for merging a patch<br>
<br>
The intent of this proposed change is to get additional test coverage<br>
and reduce the number of regressions that can sneak by. Jenkins would<br>
continue to provide +1s for all submitted changes that pass basic<br>
smoke tests. An additional +2 would be necessary from somebody who<br>
tests the patch. Providing a +2 for Verified would be semantically<br>
similar to adding a Tested-by: tag.<br>
</blockquote>
I have a basic doubt here.. How is +2 verified different than +1<br>
verified, which is currently provided by either the author or someone<br>
else or both. I assume that the Jenkins +1 verified is not the only<br>
thing that is seen by the maintainer before merging the patch, he/she<br>
should be looking at +1 verified from the author or someother person and<br>
take the decision accordingly during merge.<br>
<br>
</blockquote>
<br>
That is not the work flow model we follow currently. Authors and<br>
testers do not provide +1 verified usually and patches do get accepted<br>
with +1 verified from Jenkins. The necessary condition today for<br>
accepting a patch is +2 Code Review and +1 Verified. With the proposed<br>
change it would become +2 Code Review and +2 Verified. This change<br>
would mean that we will not merge patches even accidentally when it<br>
has been acked by Jenkins only.<br>
</blockquote>
<br>
Hmm, that would be different than the way other projects ( eg. vdsm,<br>
ovirt) use +1 verified. Wouldn't that cause confusion for people coming<br>
from different gerrit project ?<br>
</blockquote>
<br></div></div>
There are other projects which use +2 Verified too. One way or the other there are bound to be confusions. This can be handled by detailing the workflow clearly in our development-process document.<div class="im"><br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
What happens if the user / author / tester verifying the patch gives a<br>
+1 ( thinking +2 is for priviledged/maintainer ) , the workflow will<br>
still break.<br>
<br>
</blockquote>
<br></div>
It will be the maintainers' and authors' responsibility to educate such users and testers. Over a period of time we will reach a point where education would not be necessary. Till then, good documentation of this workflow and user education should provide us adequate mitigation.<br>
<br></blockquote><div><br></div><div>Another less disruptive approach is to reconfigure jenkins to give -1 verified on test failure and 0 verified on success (but still make a "passed" comment). The +1 verified should come outside of jenkins.</div>
<div><br></div><div>Avati</div></div>