Re: WG ACTION: 2 weeks to discuss [LL62] Do not space probes randomly

>If that host just happens to be the one that
>has the same address as the one being probed by the packet that is
>lost (on reception, not on the wire) every time, then the protocol
>fails.

Complete nonsense. Of course it doesn't fail. All this posturing ignores 
the fact that WE'VE BEEN SHIPPING PRODUCTS FOR SIX YEARS. This claimed 
failure is pure invention. In the vanishingly unlikely event of repeating 
identical packet loss, the conflict is subsequently detected and resolved 
a few seconds later when subsequent ARP packets are sent, and the hosts 
then pick new addresses and immediately re-probe.

The real point is that increased randomness buys us nothing worthwhile, 
and has a cost, in the form of increased variability in total address 
acquisition time, which which impacts the layers below and above:

Below: When a switch vendor asks, "How quickly do I have to bring port on 
line to be sure I don't miss all the ARP probes?" that answer becomes 
*shorter* than it would otherwise be. This tighter constraint may make 
the switch design harder.

Above: When an application writer asks, "How long should I expect address 
acquisition to take, before concluding that there may be some unusual 
error condition?" that answer becomes *longer* than it would otherwise 
be. This longer timeout results in degraded user experience.

These costs are paid by *every* host using IPv4LL, *every* time it has to 
acquire an address.

And for what?

To protect against an event that never happens, and wouldn't really 
matter even if it did.

There's a famous old quote from Antoine de Saint-Exupery which applies to 
protocol design:

<http://dictionary.reference.com/search?q=elegant>

>A designer knows he has achieved perfection not when there is
>nothing left to add, but when there is nothing left to take away.

I'm not trying to propose any radical new things here. I thought draft-07 
was pretty good. My concerns are all the weird stuff that's appeared 
since then that no one can really justify, except with answers like, "I 
think someone else wanted it."

Stuart Cheshire <[email protected]>
 * Wizard Without Portfolio, Apple Computer, Inc.
 * www.stuartcheshire.org