On 16/12/14 04:57, Noah Misch wrote:
>> But that doesn't mean we should be turning anyone away. We should not.
>
> +1. Some of the best reviews I've seen are ones where the reviewer expressed
> doubts about the review's quality, so don't let such doubts keep you from
> participating. Every defect you catch saves a committer time; a review that
> finds 3 of the 10 defects in a patch is still a big help. Some patch
> submissions waste the community's time, but it's almost impossible to waste
> the community's time by posting a patch review.
>
> Confusion may have arisen due to statements that we need more expert
> reviewers, which is also true. (When an expert writes a sufficiently-complex
> patch, it's important that a second expert examine the patch at some point.)
> If you're a novice reviewer, your reviews do help to solve that problem by
> reducing the workload on expert reviewers.
I should add that by reducing the barrier of entry for patch review,
there is definitely potential to find common errors before a patch gets
analysed in detail by a committer.
For example I may not know a lot about certain PostgreSQL subsystems but
from a decent C coding background I can pick out certain suspicious
things in patches, e.g. potential overflows, pointer arithmetic bugs,
spelling mistakes/incorrect comments etc. and flag them to the original
submitter to double-check.
ATB,
Mark.
