Hi Orie,
At 06:43 AM 07-04-2025, Orie wrote:
Sorry if my post was derailing, I worked on Decentralized
Identifiers previously.
I read your post as someone trying to express something (I don't see
anything wrong with that). I ran it through a LLM a few minutes ago
and I found out within seconds that ITU-T Study Group 17 published
X.1281 in 2024.
Many of the solutions proposed by DIDs relied on blockchain, or
merkle trees, but where updates were under the control of specific keys.
In these cases, control was split up:
- private key controllers (update control)
- registry operators (distribution control)
Governments have typically controlled the gateways to these
networks, although it's possible for them to control the network
itself with enough compute and incentive.
Depending on the threat model, you might need 3rd party
verifications to trust a name associated with controlled
identifiers, or to accept an update to the networks as a registry operator.
Observers of these artifacts build fancy graphs, where deadnames
live forever, along with other known aliases, and correlated
identifiers based on metadata analysis.
... As a practical matter, we don't control how other people
identify us, and we should emit identifiers accordingly, biometric
passport controls are a nice reminder of this... I regret renewing
my passport with a beard.
Thanks for explaining the above.
The most direct solution to the original problem seems to be for
publishers to allow authenticated authors to update their display
name, in formats that support that, without destroying the original
information.
That way the publisher can reserve the right to reject names which
are specifically chosen to discredit, attack or disrupt.
Identifying authors, or giving them control over their identifiers
is a distraction from the original problem...
If the publisher won't take the update, proving you control the
identifier does not matter.
That's a good point.
Regards,
S. Moonesamy
