W3C home > Mailing lists > Public > public-webrtc@w3.org > June 2013

Re: Signaling without a server

From: <piranna@gmail.com>
Date: Wed, 12 Jun 2013 10:25:01 +0200
Message-ID: <CAKfGGh228-MtXZcVMa2ftJ0r2oBHFpccqi3LNLqBh9yszb_BsA@mail.gmail.com>
To: Cullen Jennings <fluffy@iii.ca>
Cc: public-webrtc <public-webrtc@w3.org>
> Assuming that static web content knows about a STUN or TURN server and
has some way to know what peer IP/port it wants to connect to, I can't
think of any problem with doing this. Did you find some specific problem in
the spec?
Spec does't allow to just set the remote IP/port, just call to
setRemoteSession() giving it the other endpoint SDP. I didn't test about
making by hand a SDP with just this data, it would work but I don't think
so since there're session IDs and timestamps and so (I'm not sure, I'm not
a networks guy and I just know the basics about how SDPs works that I have
been able to learn thanks to WebRTC spec, Wikipedia, Google searching and
reading this mail list during the last year... I've not work directly with
it except to just be able too extract the session ID).

Also SDPs has info about how to traverse NATs thanks to the STUN servers so
without it only computers directly connected to internet (not local LANs
nor firewalls) or the ones with configured port mapping would work thks
way, but it's an acceptable limitation (it's how things are currently done
for native applications anyway...).
Received on Wednesday, 12 June 2013 08:25:28 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:17:44 UTC