#96: Metadata: Should it be dynamic?

 Logs could advertise parts of the metadata (for example, the MMD) on a
 well-known location, sign it and have clients fetch it prior to other
 requests from the log.
 This means certain parameters could change frequently and logs would know
 exactly how clients should behave when metadata changes.
 The downside is that we would have to consider each piece of metadata and
 the implications of allowing it to change dynamically. For example, if the
 log is allowed to dynamically change the MMD, what happens to STHs issued
 "at the seams" ? Does the old MMD apply for them? the new one?

 *not* advertising parts of the metadata directly by the logs themselves
 mean logs have no way of knowing which version of the metadata clients
 have (as it would depend on external update mechanisms such as browser
 updates).

 A way to prevent this uncertainty is to declare the metadata immutable -
 i.e. a log roll-over would be necessary to change any of the log's
 parameters.

-- 
-------------------------+-------------------------------------------------
 Reporter:               |      Owner:  draft-ietf-trans-
  [email protected]       |  [email protected]
     Type:  defect       |     Status:  new
 Priority:  major        |  Milestone:
Component:  rfc6962-bis  |    Version:
 Severity:  -            |   Keywords:
-------------------------+-------------------------------------------------

Ticket URL: <http://trac.tools.ietf.org/wg/trans/trac/ticket/96>
trans <http://tools.ietf.org/trans/>

_______________________________________________
Trans mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/trans

Reply via email to