Jari,
I oppose changes made to the document in the last round as stated
below. If they remain, I would urge publication as Informational and
not BCP:
In particular, architectural decisions, including which existing
technology is re-used, may significantly impact the vulnerability of
a protocol to PM. Those developing IETF specifications therefore
need to consider mitigating PM when making these architectural
decisions and be prepared to justify their decisions. Getting
adequate, early review of architectural decisions including whether
appropriate mitigation of PM can be made is important. Revisiting
these architectural decisions late in the process is very costly.
In the fact of a lack of common understanding regarding the threat, this
text can subject a working group to abuse and confusion. This isn't
theoretical, as it has happened in the past that working group chairs
and area directors in particular have derailed efforts, setting
standardization and deployment of useful technology back years, and this
wasn't that long ago.
Let's make this discussion concrete with a few examples: the
implications may be that a working group chair or any participant
(although chairs are in a very good position to cause damage) could
insist that DHCP not be used to carry new attributes because there is no
common understanding of the scope of remediation that will be required.
Before certain ADs roll their eyes, the discussion gets derailed as follows:
I propose the following DHCP option to configure my new frob.
<< But DHCP is transmitted in the clear. Please justify this decision.
(or worse) << and by the way here is my very heavy weight alternative
that requires a valid cert chain
It's meant for the local wire.
<< But we don't know the scope of the attack.
...
...
Nevermind, I'll just use a vendor extension. Goodbye.
Rinse and repeat with any other protocol that allows extensions.
It is fair to say that we should consider this threat at an
architectural level. It's fair (albeit a truism) that finding design
flaws earlier in the process rather than later is less costly
(ENG-101). Justification language like the above, however, is likely to
actively impede the IETF, as these sorts of things have in the past.
Eliot