Where to for Samba4? - Native IPv6

[davidholder at erion.co.uk]

Jelmer,

Absolutely! The point is that it is possible to write code using IPv6
scoket API and IPv6 addresses that is actually talking IPv4. This means
you have one code base for IPv6 and IPv4.

I am back in the office next week. What I'll try to do is send you a copy
of my talk for the Google event( when I finish it) so that you can take a
look before then. Would that be useful?

Regards,
David


> Hi Björn,
>
> Am Montag, den 10.09.2007, 13:24 +0200 schrieb Björn Jacke:
>> On 2007-09-08 at 22:00 +0200 Jelmer Vernooij sent off:
>> > Am Samstag, den 08.09.2007, 20:38 +0100 schrieb
>> davidholder at erion.co.uk:
>> > > Great to hear from you.
>> > >
>> > > Going IPv6 native does not get rid of IPv4. The IPv6 API is by
>> definition
>> > > dual stack (strange though it may sound). What we can do is make all
>> the
>> > > code IPv6 or protocol independent and still support IPv4.
>> > Just the IPv6 representation isn't sufficient because we need the IPv4
>> > addresses. For example, NBT uses a little-endian 4 byte serialization
>> of
>> > IPv4 addresses. Several of the protocols don't make sense to run over
>> > IPv6, such as NBT or WINS.
>>
>> I think what David is referring to is this:
>>
>> http://www.kame.net/newsletter/19980604/
>>
>> It might in fact be a good idea to use this API at least for those parts
>> which
>> *do* work with IPv6.
> I agree that is a good idea and am not opposing that, but I think it's a
> bad idea to provide protocols over IPv6 that can only work with IPv4
> anyway (such as NetBIOS).
>
> Cheers,
>
> Jelmer
>
> --
> Jelmer Vernooij <jelmer at samba.org> - http://samba.org/~jelmer/
> Jabber: jelmer at jabber.fsfe.org
>