[tcpm] strength of tcpsecure recommendation

Folks-

I'd like to try to be a bit systematic about figuring out how to proceed
with the tcpsecure i-d and so I have cooked a poll below.  The essential
questions on the table is the mitigations ought to be labeled as SHOULD
or MAY and whether these labels ought to be conditioned or not.

We have had some good conversation on the topic and lots of points have
been aired (how baked these solutions are, IPR, cost v. benefit, etc.).
So, I am asking that the answers to the following questions be succinct
and not rehash a bunch of previous argument.  If you feel you have a
**new** point then that is clearly fine.

I am trying to get a good read of the WG here and so even if you have
clearly aired opinions before, please take the minute and answer these
questions.

(Note that SHOULD and MAY are clearly not the only choices.  If you want
to vote for MUST or MUST NOT then that is fine, even though they are not
listed as choices below.  If you do, please then do provide some
justification.)

The questions then ...

(1) (a) Do you believe the draft should allow the "blind reset attack
        using the RST bit" mitigation described in section 3 of the
        document as a SHOULD or a MAY?

    (b) Should we apply a condition to the recommendation?  If so,
        please state the condition you'd like to see very succinctly.

(2) (a) Do you believe the draft should allow the "blind reset attack
        using the SYN bit" mitigation described in section 4 of the
        document as a SHOULD or a MAY?

    (b) Should we apply a condition to the recommendation?  If so,
        please state the condition you'd like to see very succinctly.

(3) (a) Do you believe the draft should allow the "blind data injection
        attack" mitigation described in section 5 of the document as a
        SHOULD or a MAY? 

    (b) Should we apply a condition to the recommendation?  If so,
        please state the condition you'd like to see very succinctly.

Thanks in advance for the input!

allman