Re: Re: [dhcwg] WG Last Call for DHCP draft

On Thu, 27 Jan 2005, Ted Lemon wrote:

> On Jan 27, 2005, at 2:49 PM, Vivek Kashyap wrote:
> 
> > True..I'm not convinced that a reference to every draft that may 
> > become relevant has to be referred to .. if it was an RFC I'd 
> > immediately put a reference in. The dhcp over infiniband draft is not 
> > precluding any other use.
> 
> 
> Vivek, this isn't a case of a draft that may become relevant.   The 

Yes, it is the case of a draft becoming an RFC to become relevant :). 

The 3315id-v4 draft provides a method that is going to be applicable 
to any DHCP implementation however it still does not void the use of 
other possible client-identifiers. The IPoIB WG discussed an issue and 
we came up with certain clarifications on the GIDs - these were included
in the draft. To illustrate the discussion further a few examples - clearly
stated as examples are provided. These do not preclude any future RFC 
from being followed and the draft states so.  Your draft when it becomes
an RFC will also not preclude RFC2131 client-identifier uses. Therefore they 
should be able to move in parallel.

I hope the above explains why a dependency appears uncalled for. 

PS. when you reply please do not snip away the contents -- I prefer to see
the full context of the response as I choose not to keep every response 
in the thread.

> draft that I suggested you reference solves precisely the problem 
> you're trying to address.   The reason I wrote the draft was to address 
> problems raised by the DNS working group chair and several area 
> directors in the RFCs that you are referencing - I didn't just make it 
> up for my own personal satisfaction.   

:)I'd hope so.. 

The language in your draft 
> performs the same function as my draft - it isn't a case of them being 
> loosely related.   I've made some suggestions for how to mitigate this 
> conflict, but ultimately the right thing to do is to simply remove the 
> conflict.

You confuse me...a while back you had accepted that I'll make it 
clearer than it is now that there are other methods to create client-ids.
I was fully convinced we had resolved the issue :).

Vivek

> 
> As to your question of whether the draft needs to be referred to, I'm 
> willing to listen to arguments that you needn't reference the draft, 
> and to look at text you propose to mitigate the problem.   However, 
> it's still true that the easiest way to satisfy my concerns would be to 
> reference the draft that I've advised you to reference.
> 
> You have to decide whether it's quicker to do as I ask and wait for my 
> draft to make RFC, or to try to avoid referencing the draft.   It's not 
> at all the case that I have final say in whether your draft advances.   
> Do be aware, though, that there are two reasons why I've advised you to 
> reference my draft.   One of them is that I think it's the right thing 
> to do, and will lead to the best outcome.   The second is that if the 
> draft passes the working group, I think following my advice about these 
> specific concerns is the best way to get it through the IESG.   I may 
> be completely mistaken about that, of course, but that's why I've given 
> you this advice.
> 
> 
> _______________________________________________
> IPoverIB mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/ipoverib
> 
>