Re: [hybi] frame length encoding

["Shelby Moore" <[e-mail]>]

> On Sun, Aug 22, 2010 at 6:58 PM, Shelby Moore <[e-mail]> wrote:
>
>> That forces the receiver to check for the larger size.  See my prior
>> post
>> on the multi-core issue, that it would be more efficient in that case if
>> the sender honored the maximum.
>>
>> Why can't the sender honor the agreed maximum size?
>>
>
> The receiver will still have to check, because the sender might be an
> attacker hoping to take advantage of a receiver that only checks the first
> 2
> bytes of the length while firewalls/etc would have filtered it had it been
> correct.  Getting different entities to make different decisions about
> some
> framing of data is a fundamental attack tactic.

Isn't that incorrect logic?

If the spec says that the maximum is the maximum, then firewall should
respect the spec.

Any sender can send any thing any time.  That is not relevant here.