Re: [rtcweb] Asking TLS for help with media isolation

[Paul Kyzivat <[e-mail]>]

I'm confused by this, in at least a couple of respects:

- IIUC this isolation is a browser thing. If the other end doesn't
   have a browser then what does it mean?

- Isn't the isolation supposed to be on a per-stream basis?
   (Or is it really on a peer-connection basis.) If it is per-stream
   then how does signaling it on a per-TLS basis work?

- The use of DTLS to do SRTP keying is a sort of hack, since it
   isn't about the DTLS payload at all. I would think that whatever
   new you introduce into TLS for this would most naturally apply
   to the DTLS payload (the data channels).

	Thanks,
	Paul

On 4/3/14 8:14 PM, Martin Thomson wrote:
> As I described briefly at the last meeting, ensuring that media is
> isolated from the application or web site is a key part of addressing
> our security goals.
>
> The key part of that is making sure that any media that is isolated on
> the sending side of RTCPeerConnection remains isolated when it reaches
> the other side.
>
> As I noted, this is also necessary in order to ensure the integrity of
> the same origin model.  In that model, cross origin media is required
> to be inaccessible to content, and as it stands RTCPeerConnection
> could be used to work around those restrictions (implementations can
> implement other protections, as Firefox already does).
>
> The alternatives as I see them (and I hope that this is sufficiently
> exhaustive) are:
>
>   1. ask the TLS working group for a TLS-based solution
>   2. build something into the session signaling (i.e., new SDP bits)
>   3. give up on the idea
>
> I prefer 1 for reasons already outlined.
>
> I propose that we formally request the TLS working group recommend or
> define a mechanism whereby we can signal in the TLS handshake that the
> session contents are to be protected from access (where the
> description of that protection may need to our responsibility).
>
> I have pointed to draft-thomson-tls-acp as a potential solution here,
> but others have noted that ALPN tokens could be used.  I expect that
> TLS are capable of making the right decision here.
>
> _______________________________________________
> rtcweb mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/rtcweb
>