Re: [boost] [process] [sorting] [singularity] [compute] [others] Who needs review manager?

On Sep 13, 2014, at 3:28 PM, "Edward Diener" wrote:

Using Robert Ramey' analogy this would be the same as a judge in a court case also serving as a juryman and casting his vote. I do not believe that should ever be the situation either in a law court or during a Boost library review.

Of course, but this does not mean that the review manager should be reviewing the library. A good review manager's responsibility is to be as impartial as he can be in deciding, based on the reviews and his knowledge of the library functionality, whether it should be accepted or not.

As a former review manager, I did review the code and give feedback to the author (privately), along with reading the code while reading reviews. This seems completely aligned with facilitating a smooth review period with a positive result at the end (by positive, here, I mean positive for Boost while being constructive to the author). However, I didn't see the need to submit a formal review and vote myself, since it doesn't fit with the spirit of the process as I see it.