----- Original Message -----
From: "Moti Asayag" <masayag(a)redhat.com>
To: "Eli Mesika" <emesika(a)redhat.com>
Sent: Friday, February 22, 2013 9:02:08 AM
Subject: Re: [Engine-devel] Changing Gerrit -1 message
On 02/22/2013 12:53 AM, Eli Mesika wrote:
> ----- Original Message -----
>> From: "Dave Neary" <dneary(a)redhat.com>
>> To: engine-devel(a)ovirt.org
>> Sent: Thursday, February 21, 2013 11:21:53 PM
>> Subject: Re: [Engine-devel] Changing Gerrit -1 message
>> On 02/20/2013 10:32 AM, Itamar Heim wrote:
>>> On 19/02/2013 12:06, Laszlo Hornyak wrote:
>>>> I agree with that.
>>>> for the - messages this in my opinion would be both more clear
>>>> -1: In my opinion it needs work.
>>> how about
>>> "-1: Please review my comments"
Seems that we're in agreement on this one....
>> Sounds great.
>>>> -2: I disagree.
> I prefer -2 : Do not submit!
> -1 should be used whenever the code is OK but can be done better or
> has a missing part (tests for example)
> -2 should be used when the code does not work, has a serious bug
> (possible NPE for example) , break the build
+1 for the -2 which to my opinion should indicate any serious flaw in
the design, even post the design discussion. I think that this should
follow with a discussion on engine-devel about the required change in
the design instead of debating over the gerrit since it is not
-2 should be given on rare occasions, as it is so blunt and meaningful.
Remember that -1 is good enough to block a patch.
So if you see an issue which may be fixed, give it -1.
If you see something which is illegal (violate project license) or
contradicts the project foundations, give it -2. This is how rare -2
In this view I'd rephrase "-2" to: "This patch can not be
Obviously this should be followed by a detailed explanation by the reviewer.
>>> "-2: Please reconsider"
>> I am trying to think under which circumstances people give a -2.
>> something like "We have discussed this feature, and I disagree
>> is good for the project." Basically, I don't think that a
>> should ever see a -2, unless they specifically disagree with the
>> maintainer, and insist that they are right, to the point of
>> submitting patches.