https://datatracker.ietf.org/rg/hrpc/about/
IETF 117, San Francisco
2023-07-26, 1300-1500 local time
Chairs: Sofia Celi and Mallory Knodel
Notetakers: Shivan Sahib, Ryo Yanagida, (add yourself here!)
Mallory: Wrt the factors impacting digital repression; does make it
further damaging? [will revisit]
Steven: govt has a number of ways to exert oppression: violence
being one. Digital repression is a simpler alternative (e.g. Turkey and
Erdogan). Substitution effect.
Vittorio Bertola: If I take any of the definition of the digital
opression, and show this to people in the 'west' that are being
'censored' by the platforms like facebook (e.g. antivaxxers), they'll
say that's what's happening to them.
Should we also look at what's happening in the 'west'? Is that also
digital repression?
Steven: I'm not sympathic about putting those censorship on platform
inline with the digital repression we are seeing in more oppressive
regime. The factors/measures I used here are well regarded model from
credible institutions like Freedom House.
Mirija Kühlewind: Do you have any knowledge of how much these regime
invest in controlling information?
Steven: It really depends and have some data. I looked in to
spyware, and they can be 100 mils of dollars etc. Some financial
disclosure showed a new tender from the Indian government related to a
spywhere and that's aboout 140 mil.
Tobias Fiebig: like to challenge point that it's easy to
differentiate between digital repression and legitimate misinformation
squashing or illegal content blocking. From a tech pov, we allow banning
of content based on the rule of law.
Steven: In democratic system, the representatives can be voted out
etc. to influence the policy of the government. It's not perfect, and
have flaws, but have a measures to correct/fix/influence from
population. Compared to that, in non-democratic system, that's just not
possible and there are no accountability.
Dan Harkins: Have you looked into the social scoring/credit systems
like seen in China? In US/Canada people can be denied banking etc. too,
albeit it's a cruder system though.
Steven: The importance of the social score is exaggerated. Trend is
concerning though in both China and "Western" democracies.
Nick Doty: Do you have any evidence in the shift of trend in the
specific tactics, like some particular tactics are on the rise etc.
Steven: It's difficult to derive. Are the efforts rising because
there are more people on a platform, or [...] This is hard to
quantify, so we end up having to aggregate, estiate.
Matt Mathis: How did youd define 50%? We might see 'different' 50%
John: 50% of publicly available IP space. You can't have more than
50% that is 'completely different'
Let's take it offline.
Nikita Borisov: When you say public IP addresses, you are talking
about those actually allocated by IANA?
John: This will be a long conversation. Many are ephemeral, and
there are ways to handle this. Depends.
Dave Plonka: ICMP is an interesting way of measuring this but not
everyone responds to pings, which you would label as fragmented. What is
the best way of measuring this passively instead?
John: Not everyone responds to active probes. I'm working on a
passive mechanism but that's hard. It is vital. This talk focuses on
active and does give a good coverage for now. Will work on passive
methods. Next paper!
Q Misell: So you have 6 vantage points, have you looked into RIPE
atlas?
John: See paper; it's not fast enough and we don't want to abuse it
[not put too much load on it].
Q Misell: You expect the result to be the same?
John: Their results confirmed our result.
Tobias: Are you looking to increase resolution of the measurements
via greater number of probe points?
John: There are a lot in this topic. We did use RIPE atlas to
estimate some of the measurements. So far, we have a pretty good
resolution but we could certainly improve with more probes.
Nick Merrill: I like this crisp definition. What is the 'canonical'
internet? Is there one?
John: We thought there would be one Internet. Perhaps there are
perspectives where we want "more than one Internet" (sovereign
internets). Important result of the definition is, that "one" internet
is not inevitable. If we lose it, we lose it.
Mallory: there was a talk last time at HRPC that showed lots of leaks
are happening.
Tobias Fiebig: Wrt Email; not as encryptes as we might think.
Opportunitistic TLS etc.
Philip Hallam-Baker: Good work; the patterns we are seeing is 'hack,
frame, and leak' pattern.
Snowden showed that governments do hack. If we are going to protect
democracy, we need stronger, and better encryption.
Stephen: Agreed
Andrew Campling: About RFC7256; perhaps the community paid too much
attention on government doing bad thing, while they ignored what
corporations do to people or SNS does. Are we enabling these sectors to
continue doing 'bad stuff'.
Stephen: I don't fully agree. It is much easier to react against the
government and much harder to go against the coorporate, which may well
be your sponsor/employer.
Alissa Cooper: I'd object the statement that the community put blind
eye to the cooporates doing bad things. People who have been working on
it has continued working on it. [this is for Andrew]
Stephen: As a community, some of it perhaps might be due and worth
taking.
Vittorio Bertola: What's the next challenge? To me, it seems they
are different now. When IETF started, the world was in a simpler world
with simpler model w/ one server, one client — it's much more
complicated now.
Stephen: It is fair that centralisation is much more pressing matter
now. But here, we can probably get more stuff regaridng encryption done
more so than those other aspects.
Mallory: Perhaps more 'narrow case' might still be in scope.
Niels ten Oever: We are not setting what IETF should do, but provide
guidelines for how to write human rights considerations.
Eric Resco: But the document says "should do xyz" in various places.
Specifically referring to protocol development/definition.
Mallory: Agreed, some parts are quite direct and clear. However, in
general, I don't completely agree that is the case in all aspects. Not
everything is 'right' or 'wrong' as it is very complicated.
Eric: It seems to render this document less helpful if it isn't
actually providing me if it didn't come with any judgement
Mallory: The text may be useful but may still be unsatisfactory. But
that might be the nature of this
Colin Perkins: There are some differences in the opinions on this
with different expectations on the draft and also the interpretations of
the text itself. I'd encourage people to be more aware of differences of
others.
Joey: Usefull is always debatable but there is a clear distinctions
between guidelines and rules.
Sofia Celi: Comments came in quite late
Colin: We don't dismiss comments that came in late about correctness
simply due to the lateness. Some clearly seems to have a point.
Gurshabad: I don't think this is an issue of correctness; we don't
have the time now but we'll continue working on this.
Mallory: Thanks all for the contributions, we'll continue working on
the list and github.