Hi Julius,
Sandrine is away on vacation so I thought I would follow up as we had a chat in one of our team meetings about the point below.
> On 07/07/2020, 23:05, "TF-A on behalf of Julius Werner via TF-A" <tf-a-bounces@lists.trustedfirmware.org on behalf of tf-a@lists.trustedfirmware.org> wrote:
>> IOW, you're fine with both options but you'd prefer to have different >> people setting MR+1 and COR+1 (or CR+1 in the special cases mentioned >> above), right? Just want to double-check I understood your position.
>Right.
So when we had +1/+2 CR before these changes we endeavoured to ensure that these were done by different people and if a maintainer gave a +1 they resisted also giving a +2 as we recognise having multiple eyes is always advantageous. So the general feeling was that this should be the recommended best practice. However, the feeling was also that this specific recommendation should not be strictly enforced by gerrit checking as on occasions it is possible to give both on simple patches and maintainers and code owners in this case can be trusted to make the correct choice.
The enforcing of them in Gerrit should be viewed as an aid to try and avoid mistakes rather than strict control for as you say we don't expect people to intentionally break the rules. Perhaps a warning in this case can be made to remind folks what's best practice here.
For now though as mentioned as we practice the new rules and recommendations we will use an honour system rather than strict enforcement along with updating the documented rules and recomendations.
Thanks
Joanna