[OpenID] Comment on new Draft host-meta

[Peter Williams]

So I enjoyed the paper, but found myself glossing over the second half. It's
very mid 90s. The author has not moved on from the conceptions and is still
using that period's framing of the issues. It didn't gel with the ebb of
today's web current (manifest in Google's conceptions for the openid-enabled
cloud); which are more evolutionary than revolutionary.

At heart, it assume the agent hypothesis (that a PC can be trusted to be an
intermediator between human space and machine space, and the PC is not
subverted by politics). I assume my CPU core is a latent spying device with
auto-correlation properties that identify me, the current user.