PING^2: resubmitted IRA improvement patches

[Richard Kenner]

> I think that the SC somehow monopolized a decision on reviewers and
> maintainers.  The decision should be made on a technical expertise base
> which is out of SC competence (it could be if most SC members have
> best expertise in GCC code base).

In my opinion, the most important part of code review is NOT to see if
the code will work.  The person who wrote the code is supposed to have
tested it, so the reviewer is entitled to ASSUME that it works.  I see
the review process as more of POLICY: is the code properly formatted
and documented, is it something that's of benefit to the project, is
it the "proper" way to do something, etc.  Some of that required
detailed "technical expertese", but most of it just requires experience
with the project.