Versions the key exchange %plea and %boon. Also fixes a bug in key
generation where the vane and care were not added to the ames key
generation request.
Unifies single and multiparty encrypted remote scry to universally use
spat/stab for establishing an atom <-> path bijection, use crub:crypto
and fixes case handling for multiparty remote scry. Moves key
identifier into the spur, retains the case as the version tag, in
keeping with the singleparty encrypted remote scry. Also fixes path
handling with prepending ship/life/rift to the spur in the correct
location.
We remove the local<->eauth switching tabs entirely. The only way to see
the eauth login page is if the url contains an "eauth" query parameter.
(Entering the local ship name into the eauth form still brings you to
the normal login screen.)
The "proceed as guest" blurp no longer shows the guest identity, and we
clarify the language to indicate that it is not guaranteed to result in
access to content.
Both of those changes help simplify the look and feel of the login page.
Various users had complained that it was confusing, given that logging
into your own ship is still the common case, and not everyone knows what
"euauth" is.
For applications redirecting to the login page, it is still recommended
to add the ?eauth GET parameter if non-local sessions are supported. For
applications that don't it should now be more obvious that logging in
through eauth won't do anything (since it's not presented as an option).
We also remove some debugging prints that somehow remained in the
javascript.
We remove the local<->eauth switching tabs entirely. The only way to see
the eauth login page is if the url contains an "eauth" query parameter.
(Entering the local ship name into the eauth form still brings you to
the normal login screen.)
The "proceed as guest" blurp no longer shows the guest identity, and we
clarify the language to indicate that it is not guaranteed to result in
access to content.
Both of those changes help simplify the look and feel of the login page.
Various users had complained that it was confusing, given that logging
into your own ship is still the common case, and not everyone knows what
"euauth" is.
For applications redirecting to the login page, it is still recommended
to add the ?eauth GET parameter if non-local sessions are supported. For
applications that don't it should now be more obvious that logging in
through eauth won't do anything (since it's not presented as an option).
We also remove some debugging prints that somehow remained in the
javascript.
