Re: [Dcrup] Is there anything this WG wants to do not yet in draft-ietf-dcrup-dkim-crypto-01 ?

[Scott Kitterman <[e-mail]>]

On Tuesday, June 20, 2017 04:01:18 PM John R. Levine wrote:
> > I think it's still not clear where the group lies on how dead we should
> > kill rsa-sha1.  I think the WG chairs are going to have to evaluate the
> > consensus and let us know.  Once that's done, converging on the correct
> > wording should be ~easy (whether it ends up in my draft or John's).
> 
> Given that we are not the Network Police, I don't see this as a very
> meaningful questions We can tell people not to sign with sha-1 or to
> verify sha-1 signatures, but we can't threaten them with penalties if they
> ignore us.
> 
> In practice, the known problems with sha-1 aren't yet very relevant to
> short lived signatures like DKIM's since the cost of finding a collision
> remains high, so signers can't be bothered to upgrade their signers.  I
> looked at the last 1000 signed messages I got from non-spam bulkish
> mailers, and 354 of them still had sha1 signatures.  They can't be
> bothered to upgrade, and the IETF has no leverage.
> 
> At some point I expect one of the big gorillas will decide to stop
> accepting sha-1 and after a week of panic in the bulk mail community,
> they'll all patch their signers and we'll never see another sha-1
> signature again.

Right, and when the panic happens (which is, I agree, when this will change 
for real) I'd like for the engineer getting beat up by his boss to be able to 
say in response to the "Argh! What should be do" question from the PHB [1] "We 
need to upgrade our DKIM like it says in RFC XXXX" [2].  I think what you've 
proposed covers that.

Scott K


[1] The pointy haired boss (see Dilbert), not the other one.
[2] Followed by a mostly silent "Like I've been telling you for a year."