[Netconf] Questions about draft-ietf-netconf-tls-04.txt

Hello,

Regarding NETCONF over TLS, I have two questions.  If these questions 
have been answered already, would someone kindly point me to relevant 
information?

RFC 4742 (NETCONF over SSH, Section 3.1) specifies a "framing sequence"
to be inserted after each NETCONF message.  Why is this framing sequence
not needed or specified for NETCONF over TLS?  It would seem that in either
case, the transport layer (SSH or SSL) is simply a "data pipe", and that 
if a framing sequence is needed for SSH, it would also be needed for TLS.  
Making the framing sequence mandatory over SSH but not over TLS requires 
two slightly different software implementations.

Second, when implementing a NETCONF server over TLS, how is a user identity 
derived for the purposes of data view access control?  Although not specif-
ically mentioned in RFC 4742, I believe such a user identity can be obtained 
from the process environment of the NETCONF server SSH subystem (on open
systems, at least), based upon the username specified during the launch of 
the SSH client.

Thanks in advance for any/all information!

Regards,
--Alan

_______________________________________________
Netconf mailing list
Netconf@ietf.org
https://www.ietf.org/mailman/listinfo/netconf