ietf
[Top] [All Lists]

Re: Best Effort Key Management (was Re: [saag] Last Call: <draft-dukhovni-opportunistic-security-01.txt>

2014-08-06 10:32:04

Dave,

I'm not sure the below is of that much use to be honest,
and will as I said be going over the full thread when I
can, in part to try debug whatever it is that causes you
to have such a different opinion, but in the meantime...

On 06/08/14 15:16, Dave Crocker wrote:
On 8/6/2014 7:00 AM, Stephen Farrell wrote:
However I *really* do not think we'd be wise to re-start the work
looking for a new term or a new meaning so I won't comment on
your suggestions along those lines.


Stephen,

That type of response represents what is called a 'sunk cost' error. It
entails continuing with a problematic choice because of its time in
service, rather than switching to a better choice, in the false belief
that the problematic issues can be resolved.

The problems with the term 'opportunistic security' are multiple,
serious and inherent.

I fully disagree with the above assertions. There are quite a
few people who find the term just fine and Viktor's draft fine
enough. (Note: I'm not now asserting that there's an IETF nor a WG
consensus, but we did test this on the saag list and should not
try force everyone to repeat that discussion.) There are it
seems a few including you who find things problematic. We need
to figure that out, but before doing that I am not willing to just
accept on the basis of your assertion that you are correct and
all those with other opinions are wrong.

My preference for now would be to park this aspect for a few days
so I can go get back up to speed on the full thread if that's ok.
(While at the same time knowing that suggesting that is likely to
ignite even more discussion;-)

Cheers,
S.




What happens with sunk cost errors is that the folks working closely on
the topic become attached to it and are frankly unrealistic about how
deep and intractable the problems are.

In the case of a vocabulary exercise, as we are having here, folks in
the core effort get comfortable with some sort of shared 'sense' of
things and miss the fact that even that vague sense will be missing when
the rest of the world uses the term.  Instead, the rest of the world
will spontaneously and repeatedly invent whatever suits it.

Your citing a scope of use as IETF-ish folk represents the problem.  We
must not be an enclave.

So what we are going to get is a term that has no obvious and clear
meaning.

d/


<Prev in Thread] Current Thread [Next in Thread>