Re: [dhcwg] behavior on lifetime expiration (Re: comments on draft-ietf-dhc-lifetime-01.txt)

On Sat, Sep 04, 2004 at 12:33:37PM +0900, JINMEI Tatuya / ?$B?@L@C#:H wrote:
> >>>>> On Fri, 3 Sep 2004 10:34:50 +0200, 
> >>>>> Stig Venaas <Stig.Venaas@uninett.no> said:
> 
> > So how about something like this:
> 
> > When a client receives a Reply to an Information-Request that
> > contains configuration information (i.e., does not contain a
> > Status Code option), it SHOULD install that new configuration
> > information after removing any previously received configuration
> > information.  There may be reasons not to always do this.  One
> > example might be when client has indication that it has moved to
> > a new link.  How a client copes with movement is outside the scope
> > of this document.
> 
> > I'm not not sure if it should be "SHOULD" or "should" there. At
> > least I don't see this as part of the protocol. In my opinion
> > the idea is just to give a general recommendation or some guidance
> > to implementors.
> 
> The proposed text basically looks fine.  But I'd be a bit more
> specific that the above description means if the new configuration
> information does not have any update on a particular instance of the
> old information (e.g., DNS server list) the client will simply remove
> the old information.

Note that I say "after removing any previously received configuration
information". To me at least, that says that ALL the previously
received data is removed when receiving new information. It doesn't in
any way say that this is on a per option basis. E.g. if you previously
received options A, B and C, and you in the new message receive only a
new option D, then you should still remove previous information (A, B
and C).

But, if you still think this is not clear, I will try to improve it. If
it's not clear to you, then it may not be clear to others either.

Stig

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg