Hello,
>>>>> On Thu, 9 Jan 2003 09:44:18 -0500,
>>>>> "Bound, Jim" <[EMAIL PROTECTED]> said:
> How are you and I hope well.
Thanks, I'm fine.
Perhaps my previous response was too simple, and I should have been
more careful. I said:
>>>>> On Thu, 09 Jan 2003 21:31:16 +0900,
>>>>> JINMEI Tatuya <[EMAIL PROTECTED]> said:
>>> However, you may have to note that the implementation does NOT support
>>> stateful address autoconfiguration via DHCPv6, NOR WILL.
>> I don't know a lot about DHCPv6, but...why not ?
> Because we don't need it.
I actually meant "some IPv6 users around me and I (personally) don't
need it," that is, I didn't mean "IPv6 users in general don't need
it." I know some other people need for the stateful address
autoconfiguration (that's why I explicitly noted that KAME does not
and will not support this part of DHCPv6.)
I believe we (here I mean "we" in general) share the same goal of
deploying IPv6, but implementation priority sometimes varies. KAME
(of course) does not have infinite implementation resource, so we
(KAME) must prioritize what spec on IPv6 we'll implement and when
we'll do it, just like other implementors.
At this moment, we (KAME) don't see the need for stateless address
autoconfiguration *around our local community*, which affects our
current status and plans in the near future. On the other hand, we
(KAME) do see the current need to implement prefix delegation, so we
prioritize to implement that part first. Please feel free to call
this view as lack of engineering design, but this is the fact around
us for now, and we're not almighty to implement all IPv6 spec at once.
I've never stated that no one needs stateful address
autoconfiguration, and have never discouraged others to implement or
use (if there's an implementation) the feature. I apologize if I did
(and perhaps it was due to my bad wording.) It would be great if
someone who has a motivation modifies our code base of DHCPv6 to add
the support of stateful address autoconfiguration, and provides BSD
(and other open source community) with the result. In fact, I once
had an offer to contribute the stateful address configuration part to
our implementation, and I appreciated the offer (unfortunately, the
offer did not make for other reasons.)
As for the relationship with the BSD community, KAME does not own it,
and KAME cannot control their policy of adoption or distribution.
Also, regarding DHCPv6, we've not even merged our current
implementation to the BSD distribution. This is partly because we
know our implementation is incomplete (due to the lack of stateful
configuration), and partly because DHCPv6 has not reached the RFC
status for a long time (I know it has finally made it.) If the BSD
community wants to implement DHCPv6 with stateful address
configuration, either based on our current implementation or not, it's
just fine. Also, if the BSD community decides to drop KAME in general
for whatever reason, it's also their choice (we'll of course be sad,
but, again, we do not control their policy.)
I hope you accept this lengthy excuse. Also, I'd apologize in advance
for not answering your comment line by line. I tried to cover the
general point you made, but if I missed something particular, please
point it out.
Thanks,
JINMEI, Tatuya
Communication Platform Lab.
Corporate R&D Center, Toshiba Corp.
[EMAIL PROTECTED]
---------------------------------------------------------------------
The IPv6 Users Mailing List
Unsubscribe by sending "unsubscribe users" to [EMAIL PROTECTED]
