opensubscriber
   Find in this group all groups
 
Unknown more information…

s : syslog@lists.ietf.org 21 November 2007 • 4:23PM -0500

[Syslog] transport-tls-11 review
by Rainer Gerhards

REPLY TO AUTHOR
 
REPLY TO GROUP




Hi all,

I reviewed tls-11 today. Some notes:

Section 1.1: shouldn't it simply refer to -protocol for terms defined
there? I think it makes it more consistent.

Section 4.2:

===
   Authentication in
   this specification means that the recipient of a certificate must
   actually validate the certificate rather than just accept a
   certificate.
===

Is this "must" intentionally in lower case? If so, is this plausible?


Section 4.3.1: typo "tranport"

Section 5.1:

===
The server MUST be implemented to support certificate and certificate
   generation,
===

I do not think it is a MUST that a server must contain code to generate
certificates. This should be left to the implementation. There is
already the requirement to use certificates, so IMHO it is not the
business of an IETF document to specify how they are provided. For
example, I  would provide a helper app with my syslog implementations,
but not include it in the core app - it doesn't belong there.


----

Other than that, I find the draft is quite acceptable.

Rainer


_______________________________________________
Syslog mailing list
Syslog@list...
https://www1.ietf.org/mailman/listinfo/syslog

Bookmark with:

Delicious   Digg   reddit   Facebook   StumbleUpon

Related Messages

opensubscriber is not affiliated with the authors of this message nor responsible for its content.