Re: [secdir] [Uri-review] End of Last Call for draft-ietf-behave-turn-uri

[Ted Hardie <[e-mail]>]

Hi Marc,

Thanks for the changes; I thought you had suggested using new
productions, rather than re-using the existing ones from the
hierarchical
URI mechanism.  Sorry if I did not reply on that--I think that would
be a good idea, but if there is rough consensus for the current approach,
I am happy to go along.

regards,

Ted Hardie


On Mon, Nov 9, 2009 at 10:29 AM, Marc Petit-Huguenin <[e-mail]> wrote:
> I just released a new version of this I-D incorporating all the modifications
> requested during Last Call:
>
> http://tools.ietf.org/rfcdiff?url2=draft-ietf-behave-turn-uri-04
>
>
> There was only one major modification in this new version, which is the
> filtering of the list of preferred TURN transport when the scheme is "turns", to
> prevent the use an UDP or TCP transport in this case.  The reference
> implementation was updated to reflect this and is available here:
>
> http://ietf.implementers.org/turn-uri-0.2.zip
>
>
> I made some proposals during the discussion that were never acknowledged, so
> here the list of them, this the modification made in the new version of the I-D:
>
> - Ted Hardie found confusing to reuse elements from the hierarchical URI syntax
> when the URI is opaque.  No more guidance was provided[1], so I just added a
> sentence explaining this.
>
> - In the same thread, Ted Hardie pointed out that the text didn't explained
> clearly that the list of preferred transports was not an input for the TURN
> parser but for the resolution algorithm.  The I-D was modified as proposed[1].
>
> - Following the secdir review, Pasi Eronen requested some additional text to
> deal with TLS.  The I-D was modified as proposed[2].
>
> - Following the security bug discovered by Margaret Wasserman, I started a
> discussion[3] on the BEHAVE mailing-list asking if it was OK to be able to use a
> TLS transport even if a "turn:" scheme was used.  There was no subsequent
> discussion on this, so the I-D now prevents to use a UDP or TCP transport if a
> "turns:" scheme is used, but does not prevent using a TLS transport if a "turn:"
> scheme is used.
>
> - Following the ops-dir review by Margaret Wasserman, I started a discussion[4]
> on the BEHAVE mailing-list for opinions on the implicit processing in the I-D.
> There was no subsequent discussion on this, so the implicit processing was not
> modified in the I-D.
>
> - The last iteration of the modifications[5] for the algorithms steps were
> integrated in the I-D.
>
>
> Here's the full changelog:
>
>   o  Improved the algorithm steps.
>   o  It is possible to use a TLS transport event if the scheme is
>      turn:.
>   o  Clarified when to stop the resolution with an error in step 2.
>   o  Added transport list filtering process.
>   o  Improved security section following sec-dir review.
>   o  Fixed nits reported by gen-art review.
>   o  Added example for remote hosting.
>   o  Removed URIs section.
>   o  Editorial modification.
>
>
> Many thanks to all the reviewers.
>
>
> [1] http://www.ietf.org/ibin/c5i?mid=6&rid=49&gid=0&k1=933&k2=49076&tid=1257785026
> [2] http://www.ietf.org/mail-archive/web/secdir/current/msg01205.html
> [3] http://www.ietf.org/mail-archive/web/behave/current/msg07289.html
> [4] http://www.ietf.org/mail-archive/web/behave/current/msg07292.html
> [5] http://www.ietf.org/mail-archive/web/behave/current/msg07314.html
>
> --
> Marc Petit-Huguenin
> Personal email: [e-mail]
> Professional email: [e-mail]
> Blog: http://blog.marc.petit-huguenin.org
> _______________________________________________
> Uri-review mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/uri-review
>