Re: [v6ops] why IPv6 EHs in the Real World

[Brian E Carpenter <[e-mail]>]

On 11/04/2015 04:29, Nick Hilliard wrote:
> On 10/04/2015 00:58, Brian E Carpenter wrote:
>> On 10/04/2015 09:07, Nick Hilliard wrote:
>>> products are designed to a budget and a performance spec.  There's no point
>>> in building a router which will inspect 65536 byte packets for EH chains if
>>> it's too expensive for anyone to buy, or if it trashes performance by
>>> passing packets with headers that big.  There is always a
>>> cost/speed/features compromise.
>>
>> Of course. But we already changed the standard to limit this problem:
>>
>> http://tools.ietf.org/html/rfc7112
> 
> a policy cannot change reality.  At best it can hope to complement it.

True. But fixing our specs and encouraging implementers to do the
right thing are the only tools the IETF has.

    Brian