Re: [babel] minor DTLS comment

Thanks Barbara, how about this slightly more pedantic option?

Babel over DTLS operates on a different port than unencrypted Babel.
All Babel over DTLS nodes MUST act as DTLS servers on a DTLS port, and MUST
listen for unencrypted Babel traffic on an unencrypted port, which MUST be
distinct from the DTLS port.  The default port for Babel over DTLS is
registered with IANA as the "babel-dtls" port (UDP port TBD), and the
unencrypted port is registered as the "babel" port (UDP port 6696).
Nodes SHOULD use these default ports.

https://github.com/jech/babel-drafts/commit/c90cc8204691254e7051405acf6a10088d42cdfd

(Dave, running Babel over alternate transport protocols sounds like
interesting future work outside the scope of this draft)

Thanks,
David

On Fri, Jan 4, 2019 at 12:26 PM Dave Taht <dave.taht@gmail.com> wrote:

> On Fri, Jan 4, 2019 at 12:07 PM STARK, BARBARA H <bs7652@att.com> wrote:
> >
> > Hmm.
> >
> > " Babel over DTLS operates on a different port than unencrypted Babel.
> >
> > All Babel over DTLS nodes MUST act as DTLS servers on a configured port,
> and
> >
> > MUST listen for unencrypted Babel traffic on a distinct configured port.”
> >
> >
> >
> > An implementation doesn’t have to allow for configuration of ports. They
> could just be hard-coded.
> >
> > I don’t like that I need the context of the first sentence to understand
> unambiguously what the second port is “distinct” from (i.e., distinct from
> the unencrypted Babel port).
> >
> >
> >
> > How about just “ Babel over DTLS MUST operate on a different port than
> unencrypted Babel.” ?
> >
>
> Sure.
>
> (I had proposed at one point that the dtls version operate over
> udp_lite or even dccp, on the same port number, but have since
> discovered osx doesn't support either)
> >
> > Barbara
> >
> >
> >
> > From: David Schinazi <dschinazi.ietf@gmail.com>
> > Sent: Friday, January 04, 2019 2:41 PM
> > To: STARK, BARBARA H <bs7652@att.com>
> > Cc: Babel at IETF <babel@ietf.org>
> > Subject: Re: [babel] minor DTLS comment
> >
> >
> >
> > Thanks for your comment, Barbara. I agree with you, and have added the
> following text:
> >
> >
> https://github.com/jech/babel-drafts/commit/a5a372a942ebc7951b2847d88123d75ab8169f2f
> >
> >
> >
> > Please let us know if you feel it addresses your comment.
> >
> >
> >
> > Thanks,
> >
> > David
> >
> >
> >
> > On Fri, Jan 4, 2019 at 5:35 AM STARK, BARBARA H <bs7652@att.com> wrote:
> >
> > Since the DTLS draft is still open for comments, I do have a small one
> about how the UDP ports are characterized.
> > The IANA assigned ports are default values, and need to be portrayed as
> such. It's allowed (or should be) for deployed instances to use other
> values. This is pretty much true of all protocols. Certainly the base babel
> protocol allows other values to be used (which is why the homenet babel
> profile mandated use of 6696).
> >
> > Maybe instead of
> >    All Babel over DTLS nodes MUST act as DTLS servers on the "babel-
> >    dtls" port (UDP port TBD), and MUST listen for traffic on the
> >    unencrypted "babel" port (UDP port 6696).
> >
> > say
> >    All Babel over DTLS nodes MUST act as DTLS servers on the "babel-
> >    dtls" port, and MUST listen for traffic on the
> >    unencrypted "babel" port.
> >    The IANA-assigned values of 6696 for the "babel" port and
> >    TBD for the "babel-dtls" port SHOULD be used.
> >
> > Barbara
> >
> >
> > _______________________________________________
> > babel mailing list
> > babel@ietf.org
> > https://www.ietf.org/mailman/listinfo/babel
> >
> > _______________________________________________
> > babel mailing list
> > babel@ietf.org
> > https://www.ietf.org/mailman/listinfo/babel
>
>
>
> --
>
> Dave Täht
> CTO, TekLibre, LLC
> http://www.teklibre.com
> Tel: 1-831-205-9740
>