[dmd-internals] To consume pull requests for trivial bugs

[Andrei Alexandrescu]

On 8/21/12 9:21 AM, Leandro Lucarella wrote:
> I think ideally somebody should review your commits, but if you've been
> offered commit rights, then it's implied you don't *need* review (nobody
> reviews Walter commits, not at least before they are pushed to the
> repo). Again, it would be ideal if somebody else review the patch, but
> I don't think it should be a blocker.
>
> But that's just *my* POV and I have no idea what does it mean for Walter
> to give commit rights (I guess is push rights in git terminology).

The way it works for us at Facebook is that every line of code must be 
reviewed by another engineer than the one who wrote it. It is working 
very well for us.

There are many positive consequences to this. If the rule were not in 
effect, I suspect the notion of "this code is mine" and "that code is 
yours" would immediately emerge. With peer review enforced it becomes 
clear that once a pull request is merged, it is owned by the entire team.


Andrei