[Cryptography] Standards Trolls: Re: Bitcoin is a disaster.

[Phillip Hallam-Baker]

On Thu, Dec 31, 2020 at 11:24 PM Deryk Makgill <makgill at makgill.ch> wrote:

> ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
> On Thursday, December 31, 2020 2:17 PM, Ray Dillinger <bear at sonic.net>
> wrote:
> ...
> And as you said, the troll army that ran disinformation campaigns and
> attacked
> node implementations and companies trying to raise the blocksize with
> letters
> to the SEC, DDoS, delistings and negative review brigades were highly
> effective.
>

Standards trolls are a very real issue in standards processes. One of the
things I find rather strange is that some people react to the mere
suggestion that nation state actors might get involved in directing
communications standards with snears and derision. Or rather I did until an
MI6 officer told me he believed an individual was using those techniques as
a nation state actor trying to disrupt the process.

That was back in 1993.

One service Snowden did perform for the field was that disclosure of the
BULLRUN program made the fact the US ran these programs public record. But
they were never the only country involved. And I really don't think it was
the people with an open NSA affiliation who were involved in the real dirty
tricks.

Another lesson from those years was that large organizations do not speak
with one voice or have a single interest. The people who were focused on
trying to liberate communications behind the iron curtain did not have the
same approach to cryptography to the people trying to maintain the decrypt
capability on poorly implemented or sabotaged schemes being used for
diplomatic or terrorist communications. Perhaps making so much noise on one
was an attempt to distract attention from the other.

IPSEC was derailed by a whispering campaign in the bars. We were assured
that we absolutely had to stay firm on end to end security with perfect
forward secrecy or nothing. And we ended up with...

Same happened when we tried to make DNSSEC deployable using 32 bit
technology after the Network Solutions acquisition. 64 bit machines were
much more expensive and without a change to the NSEC record, DNSSEC would
add roughly $50 million to the cost of ATLAS. That should have been a no
brainer. But the sneaky whispering campaign won again.

And of course it really isn't difficult to identify one of the ideological
commitments that our industry gets wrapped around its axles. Designing a
system that is 100% peer to peer with absolutely no central coordination is
really hard and ends up requiring that to be the first priority in the
design meaning that usability and everything else required for deployment
fail. People still criticize the WebPKI but it has always been much less
centralized than the Visa/Mastercard duopoly and the market leader has
changed hands three times now.

I am now adding what amounts to a shared infrastructure to the Mesh. It is
a very small amount of centralization but it is a point of centralization
and I know the whispering campaigners are going to use that as the point of
attack in the crypto community. But the alternative is DNS and while I have
no problems with people choosing to rely on DNS for service discovery I
have a really big problem with forcing people to pay the ICANN tax just to
get a permanent name for service discovery. And freeing people from the
ICANN tax requires a naming registry (but not a resolution service).

One of the lessons I learned from leftie politics in the 80s was that the
person with the biggest mouth demanding ideological purity was the person
most likely to be a government shill and that it wasn't just our government
doing the shilling.


Trolls will troll but we don't have to let them win. But we have to
understand their techniques just the same as any other attacker.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.metzdowd.com/pipermail/cryptography/attachments/20210101/41c036e0/attachment.htm>