Re: [TLS] OpSec WGLC for draft-ietf-opsec-ns-impact

[Stephen Farrell <[e-mail]>]

IMO this document is absolutely not ready to become an
RFC. It basically ignores the purpose of TLS which is
to prevent exactly the MITM attacks being espoused
here (IMO). In ignoring that, it ignores all of the bad
consequences of such MITM attacks.

As evidence:

"To proxy a TLS session, a TLS Proxy must be able to
present a valid X.509 certificate to the TLS client to appear as a valid
TLS Server;"

That assumes that an x.509 data structure created on
the fly, with no interaction with the named subject
can be "valid." That is antithetical to the intent of
x.509 and TLS.

That is such a fundamental misconstruction of x.509,
of PKI and of TLS that it alone means that the draft
is sufficiently inaccurate to be unworthy of publication
as an IETF consensus RFC.

If really necessary I'm sure many other related
problems can and will be found in the text. I hope
that waste of effort is not needed.

The basic problem here seem to me to be an effort to
justify (and perhaps promote?) existing MITM practices
without fairly or accurately describing their downsides.

As I also commented on a related document: the effort
expended on attempts to garner IETF consensus on such
text is not worthwhile.

If someone actually wrote truly objective text on the
reasons why people deploy TLS MITM attack boxes and on
the real downsides that would be different. This is
nowhere near such objective text.

My recommendation: declare this document dead. And
don't bother with the supposed BCP. Instead, try to
find someone who can write an evidence-based description
of the falsifiable pros and cons of these TLS MITM
products. And then send that to the ISE.

S.

PS: I'm sure someone will claim that being logical about
this "ignores reality" so just to get in a comment on
that ahead of time: reality includes lots of bad
engineering - if and when we want to document that, there
is an onus on us to say why it is bad and how to do better.
That would require offering evidence as to the efficacy
or otherwise of TLS MTIM boxes when compared to the
benefits of using TLS as-designed. That evidence, in my
experience, never seems to actually be offered by those
who espouse these MITM attack products.