logo       
Google Custom Search
    AddThis Social Bookmark Button
-->

Re: Reviewer heirachies ?: msg#00043

Subject: Re: Reviewer heirachies ?
On Jul 18, 2004, at 5:17 AM, Tony Bowden wrote:
On Sun, Jul 18, 2004 at 11:56:15AM +0800, John Darrington wrote:
Note that one benefit of this is that the project administrator can opt to make certain staff reveiwers, but assign a reveiwer_rank of zero. This means that
such reveiwers have no power to pass a review, but they can fail one,
if they notice some issue that escapes the attention of their superiors.

We had such a policy in one company that I worked for that used Aegis.
Shortly after we brough the review process in, it had become a major
bottleneck. Lots of people were being failed for reasons that were mostly
to do with carelessness, so we brought into a 'pre-review' phase where
junior developers could do a first-pass check of the code looking for
common problems. They had the authority to fail the change, but not the
authority to pass it.

Of course, as Aegis didn't support this natively, it meant they had to
shout across the room to a more senior developer if they were "passing"
it, so that they knew to do the proper review.

It would indeed be nice if Aegis supported doing this sort of thing.

I second that. I worked at that company with Tony and found it a nice system to allow one to keep an "everybody reviews" policy and still ease in junior and new developers.

Same idea, different venue: open source work. It would be nice if I could give most the developers the ability to review fail but not pass. They would primarily act as style and sanity guards to offload such filtering work from the real reviewers.


<Prev in Thread] Current Thread [Next in Thread>