Re: [rtcweb] Why http://www.ietf.org/id/draft-miniero-rtcweb-http-fallback-00.txt ?

Iñaki,

discussions would be much easier to follow if you just answered to the original thread, rather than opening a new one... about your question, please see inline.

On Tue, 7 Aug 2012 20:18:58 +0200
Iñaki Baz Castillo <ibc@aliax.net> wrote:

> 2012/8/7 Marc Petit-Huguenin <petithug@acm.org>:
> > Or TURN over Websocket.
> 
> Why do we need that? By using TURN over TLS we have all we need to
> avoid stupid hotel's firewalls (that provide "Internet free access"
> but just for HTTP port 80 and HTTPS port 443). It just about setting a
> TURN server listening TLS on port 443, am I right?
> 

"Do we need that" is exactly what I wanted to ask by publishing the draft. There was already some discussions in the past, someone even suggested just deploying TURN over HTTP (a bit like Marc's point on WebSockets), and TURN over TLS (or just anything on 443, for what matters) is another viable solution.

The problem IMHO is that, just as Roman pointed out in his reply, that's not always a way out. In my experience such a scenario is not only limited to those ultra-closed environments he mentioned, but that's probably debatable (who closes his network that way, probably doesn't want anything else to go through anyway).

About the "just pipe everything on 443", please refer to my reply to Hannes in the original thread, as I've answered more extensively there.

Lorenzo

> -- 
> Iñaki Baz Castillo
> <ibc@aliax.net>
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb

-- 
Lorenzo Miniero, COB

Meetecho s.r.l.
Web Conferencing and Collaboration Tools
http://www.meetecho.com