Re: [TLS] TLS Provfiles (Was: Call for consensus to remove anonymous DH)

[Stephen Farrell <[e-mail]>]

On 16/09/15 12:18, Peter Gutmann wrote:
> Stephen Farrell <[e-mail]> writes:
> 
>> We have BCP195 [1] that aims for the "general" case (for up to TLS1.2) and a
>> draft [2] (current in IESG evaluation) for the embedded case. Are those the
>> kind of thing you're after?
> 
> Sort of, but since they're not part of the TLS spec they essentially don't
> exist (I've never seen then quoted, cited, or referenced in any third-party
> standard that deals with TLS).

I'm not sure how to process that comment. You ask for X, I ask is Y==X
and your answer is that Y doesn't exist? Seems odd. ;-)

Anyway, so far 5 RFCs reference BCP195. [1] I'd say that'll grow over
time. Hopefully folks implementing will find it useful too and not only
those writing RFCs that need TLS, but I guess we'll see over time if
BCP195 got it right or not.

  [1] http://www.arkko.com/tools/allstats/citations-rfc7525.html

> 
> Another problem is that they're defined as a large collection of (often rather
> waffly) "don't do this" comments, so as a somewhat wooly blacklist rather than
> a clear whitelist.  So the BCPs aren't really a profile but more like 20-30
> pages of hand-wringing.

Feel free to collect a bunch of your own emails (hand-wringing or
not:-) and shoot those out as an I-D.

> An actual profile of TLS would be something like MUST TLS 1.1 or above, MUST
> PFS suites, MUST AES and SHA256, MUST E-then-M (and by implication what isn't
> explicitly permitted is denied).

Yes, life would be lovely if things were so simple.

S.


> 
> Peter.
>