reviewers modifying code they are reviewing

[Hynek Hanke]

Hello all,

this is not personal to anyone, we must just find the
smoothest way to do things.

1) Reviewer must not change the patch code himself,
he must always first consult the author. Otherwise code
is getting into the repository which was not reviewed by
a second person.

2) Trivial changes can be made by the reviewer, but only
if they are corrections in formatting, obvious resolutions
of merge conflicts or changes that were previously consulted
with the author on mailing list and agreed to.

The possibility to edit patches by the reviewer is there only
to minimize unnecessary sending of patches forward and
backwards, not to override the authors coding decisions.

Also providing additional comments by the reviewer, on why
the patch was accepted or what are its other consequences,
is OK and welcomed.

The proper way to edit a patch submitted by someone else is to
keep the original message, add a Review-by line, and bellow it
add your comment or a detailed explanation of your modification.

If there are no objections, I'll make this text into a new
chapter in the coordination guidelines document.

I also found it necessary to edit files as README, INSTALL,
AUTHORS, version.sh directly as a part of the release process.
I hope everyone is OK with it. I've also integrated some
last-minute patches submitted from Debian and Ubuntu BTS,
but I still plan to send them for review for inclusion in the master.

Thank you & Best regards,
Hynek Hanke