Ben,
On 2 February 2016 at 21:01, Melinda Shore <[email protected]
<mailto:[email protected]>> wrote:
Hi, all:
We're looking for feedback on ticket 121
(https://trac.tools.ietf.org/wg/trans/trac/ticket/121).
The issue is this (from Steve Kent):
"After Prague I agreed that log metadata can be supplied by
browser vendors for TLS clients hat are browsers. However, max
chain length, which was just discussed on the list (11/16-17/15),
is an example of metadata that needs to be acquired by CAs, just
like the public key for a log. The text in Section 9 needs to say
more about how non-browser log clients are expected to acquire
this data, or maybe it should say that separate docs defining
requirements for each of these types of log clients will discuss
this issue."We continue
and where it currently stands is this (from Eran Messeri):
I agree that log metadata dissemination is a topic that should be
more thoroughly discussed at some point. However, I do not think
that 6962-bis is the right place for it.
Steve has a very valid point about metadata not being specified in
a machine-readable format, nor having a part of the protocol
describing how it is disseminated.
I believe these issues should be addressed separately of 6962-bis,
because:
The topic of log metadata is one level "above" what 6962-bis
discusses.
The requirements for metadata dissemination are very different for
each type of client.
Disseminating metadata reliably is a protocol in itself, which
(once the requirements are clear) deserve its own design.
My suggestion is to punt the discussion until after 6962-bis is
out. Presumably then there would be enough different clients
implementing 6962-bis that disseminating metadata would justify
more efforts from this workgroup.
This needs further discussion - should this go into a separate
document?
If it needs documenting at all, then yes.
We continue to have dramatically different views on what needs to be
specified in
one or more RFCs so that a standard is well-specified. This is yet
another example
of our different views. If 6962-bis wants to be viewed as a spec for CT,
not just for
the CT log, then this is another example of a gap in the spec.
Steve
_______________________________________________
Trans mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/trans
