On Tue, 2013-04-30 at 15:05 -0400, David Miller wrote:

Your patch has had large, major, shortcomings every single iteration.

This is due to people including you only point out one problem per
iteration, for some reason I don't understand...

For example, you pointed out IPv6=M issue in v5, and then pointed out
disable_ipv6 issue in v6. Until you pointed them out, I even wasn't
aware of them. All the problems I have fixed, except the naming of
struct fields which I still don't agree, exist in v1, there is no
regression since v1.

As always, it is _highly_ appreciated if all of you can point as many
issues as you can in a single iteration, so that I can fix them all in
one update (if I can fix), this also saves me _a lot_ of time,
obviously.

So, I even don't understand what you are complaining about. :) I am
trying my best to address all the issues you pointed out including the
one that I don't agree at all, except the one I can't fix right now
(scope id). If anything I can improve on my side, please explicitly let
me know.

Again, I am _not_ an IPv6 expert, I am sure there are still some other
issues like disable_ipv6 that I am still not aware of, so please don't
expect me to be able to fix all bugs alone at one time, I am trying to
fix them one by one, and this is the best that I can do.

Thanks.