On Sun, Apr 8, 2012 at 1:56 PM, Gert Doering <g...@greenie.muc.de> wrote: > Hi, > > On Sun, Apr 08, 2012 at 01:43:12PM +0300, Alon Bar-Lev wrote: >> What I am asking is for people that wishes to work with me on this >> (not the ACK/NACK approach) to allow perfecting this *BEFORE* merge. > > I would actually ask for "no more cleanup changes to tun.c before 2.3" - > I have cleaned up and *tested* all non-Linux-platforms in the previous > months and everything is working well now (not looking very pretty, I'm > the first to agree on this), but *working* *now*. >
Did I wrote or implied that this is going to 2.3? > If we start to break tun.c into pieces now, this will need extensive testing > on all platforms. Since I seem to be the only developer who does full > testing of tun/tap, ipv4/ipv6 on FreeBSD, NetBSD and OpenBSD, and I do not > have much time right now (conferences and second child due in two months), > I don't see enough testing capacity for this. > > What I would really like to see is "get out 2.3 with the (big) improvements > we have right now, and start with 2.4 right after". The next massive code > refactoring should go into 2.4 > > We need more user feedback to see which bits of the IPv6 implementation are > incomplete and need to be implemented next - and without exposure, we won't > get that. The IPv6 is one of the reason the code is so messy. Please help clean it up. For example, there is no reason not to use iovec in all cases as far as I understand. > > gert > > -- > USENET is *not* the non-clickable part of WWW! > //www.muc.de/~gert/ > Gert Doering - Munich, Germany g...@greenie.muc.de > fax: +49-89-35655025 g...@net.informatik.tu-muenchen.de
