Re: WG ACTION: 2 weeks to discuss [LL63] Text could be more specific

> From: "Stuart Cheshire" <[email protected]>
>
> >If you are correct that it is fiction then I agree that this assertion 
> >could be removed. However, why wasn't it weeded out at draft 2 or 3?
> 
> It was never in the early drafts.
> 
> It seems to have appeared in draft-11, in January this year.

You are correct here - sorry.

You questioned the truth of the assertion in clause 2.11 that "early implementations of IPv4 link-local have modified the DHCP  state machine...". The answer is in the document already  - the latest draft states in appendix A.3:

   "When in INIT state, the Windows 98/98SE DHCP Client sends out a
   total of 4 DHCPDISCOVERs, with an inter-packet interval of 6
   seconds.  When no response is received after all 4 packets (24
   seconds), it will autoconfigure an address."

Windows at this point abandons its DHCP address whether or not it has a valid lease. This is one of the things which Clause 2.11 is specifically warning against and is a modifcation of the DHCP state machine to accomodate IPv4LL autoconfiguration.

The details for the Mac OS 8 and 9 also suggest that no attempt is made to use a valid existing lease if a DHCP server is not found - favouring IPv4LL.

Philip