[Ssr2-review] sections of report
kc at caida.org
Mon May 20 04:47:30 UTC 2019
i filled in all my @@ (marks where i had text),
and i've made a ton of comments on the dns ssr ws doc
and the recommendations (and a few in the q&a).
few other issues, that i'm not sure if/where to put
in the docs.
i am wondering if the final report doesn't need a whole
category about measurement and data sharing. so many of
the recommendations seem to be asking for measurement/
data sharing. but it may be that just because the perceived
need for measurement is so pervasive, making it a separate
section would be unbalanced. so i didn't try to do it.
but maybe we should consider separately collating all the
specific measurements we are asking for, and what the
specific questions we believe those measurements would answer.
because i think we do not do a great job of tying our
asks to the need/purpose/question/harm right now.
we ask for too much, and we cannot prioritize measurements
until we prioritize harms we're trying to mitgiate/prevent.
which means we need a taxonomy.. and if if we can't document
(maybe even quantify) a serious need/problem that the
measurement will inform/address, we should drop it.
in this context, i think we probably need to ask for a new
curated data section of the web site, and we should specify
urls. maybe something like.
https://data.icann.org -- for raw data sets
https://statistics.icann.org -- for derivative data sets.
and then explicit data sets we think should go in there,
and important metadata (like dates when page is updated,
which often seem missing, e.g.,
indeed, i think we should recommend that every page have
a 'content last updated' date, because trying to figure
out what's stale and what's current on the web site
seems beyond even icann staff i ask for help ;(
e.g., icann has recently posted http://www.icann.org/opendata/,
which some vague promises, no date, and a promise for updates
as progress is made.. this is not what we want. we should
be specific about each type of measurement, that every update
must include a date, type of data, expected use of data,
questions the data can answer, how to get the data, usage terms.
now to elephants in the room:
many ssr gaps are rooted in the lack of balance of
stakholders in the presumably multistakeholder process.
in particular, there is no significant representation of
public interest/consumer protection issues. it's amazing
how broad the consensus seems to be on this point.
in this context, i find the report puzzling in many places,
asking for things as if they haven't been asked for many times,
including in ssr1. are we still having the report on explicit
analysis of ssr1 implementation? because i think it's important
to state in our new set of recommendations that some of these
attempts to get more attention to consumer protection have
been going on for over a decade. and that if (when) they fail
this time, we have a backup recommendation to put in play.
that could have the word 'regulation' in it, but maybe we
figure out a way to make sure that if icann fails to achieve
what we consider needed in terms of substantial progress on
consumer protection, then icann should publish a report
stating that it failed to achieve ssr2 recommendations within
12,24,36 months. e.g., one of our recommendations could be "
icann org should seek explicit consensus with 4-5 consumer
protection agencies on how abuse should be defined and handled,
before any discussion with CPs, and publish an explicit summary
of this consensus, or exactly where the consensus got stuck,
within 12 months of this report." then, if they fail, there
is a document that asserts it, that at least advances this
groundhog day conversation to the next step?
a related recommendation could/should be an ongoing function
to track progress of these recommendations, and prove monthly
reports to SSR2, and explicitly solicits our feedback on whether the
recommendation was implemented as we intended. i.e., don't wait for
SSR3 who may have no idea what we even meant back in 2019...
i hear a growing number of others argue that we dont see real
change on consumer protection (SSR issues) until the fiscal
architecture changes. we can discuss, but one example:
"center of excellence". is a nice idea, but so long as
it answers to icann, and so long as icann answers to the
industry that solely funds it, i don't see how it will effect
change. what is its incentive structure? it is likely to just
waste many millions generating the appearance of "something
is being done". i think it's our responsibility to point out
the fundamental incentive misalignment and lack of stakeholder
balance in the ssr context, and that it is not rational to
expect anything to change while the incentives/money flows
remain as they are. which means there needs to be a center
of excellence, but it needs to be funded by several governments,
industry, as well as icann, and have statutory support for
access to data (at least for researchers to data from r&rs
within their own country.) it must not require publication
review from funding sources. it must be truly independent.
that would have a chance to escape the stalemate wrt
or if we cannot get consensus on such an idea, those of us
who agree can add it to the report as a minority view.
More information about the Ssr2-review