svn commit: r218794 - in head: . sys/netipsec
yanegomi at gmail.com
Mon Feb 21 20:15:25 UTC 2011
On Mon, Feb 21, 2011 at 11:14 AM, Doug Barton <dougb at dougbarton.us> wrote:
> On 02/21/2011 03:01, VANHULLEBUS Yvan wrote:
>> On Mon, Feb 21, 2011 at 10:21:43AM +0100, Pawel Jakub Dawidek wrote:
>>> > On Mon, Feb 21, 2011 at 09:40:25AM +0100, VANHULLEBUS Yvan wrote:
>> [RFC4868 and MFC]
>>> > You can't talk to two such peers with sysctl or without anyway. I
>>> > assume
>>> > that if someone already has tunnels configured and they work, they
>>> > work,
>>> > because the other end uses 96 bits hashes. Once he upgrades there is
>>> > no
>>> > way to get old behaviour back quickly.
>>> > You are changing on-the-wire protocol in the middle of stable branch.
>>> > Am
>>> > I alone in thinking that this is bad idea?
>> That's a good question.
>> Of other people also think it's a bad idea, I can just forget the MFC.
>> But the same problem will happen when we'll release 9.0.
>> Of course, this is easier to explain, as this will be a new branch.
> IMO RFC compliance trumps -stable here. Admittedly some small percentage of
> users will be inconvenienced, and that is unfortunate. However all users who
> start using this technology from here forward should get the full RFC
> compliant version.
> As Bjoern pointed out, we'll see more of this, not less because oddly enough
> RFC publication (like many other external factors) do not revolve around our
> release schedules. :)
Or maybe the statement:
"Conformance and stability with the rest of the conforming
machines trumps incorrectness and stability within just older versions
Is the right way to say things? Personally I think Yvan's doing
the right thing as a bug's a bug, but that's just me.
More information about the svn-src-all