Re: [rtcweb] Requiring ICE for RTC calls

[Tim Panton <[e-mail]>]

On 30 Sep 2011, at 06:15, Martin J. Dürst wrote:

> On 2011/09/30 9:40, Hadriel Kaplan wrote:
> 
>> But it's not "POTS" that's really the point - it's the "PSTN".  From the perspective of this group, the "PSTN" is everything reachable through SIP service providers: every cell/mobile phone, DSL/Cablemodem MTA, PRI trunk, SIP Enterprise trunk, POTS landline, etc.
> 
>> And way more of those than number of users and Web Browsers on the Internet.[2]
> 
> Just for the record:
> 
> This seems to be based on the assumption that a) mobile phones don't have browsers (many of them do) and b) that each user has only one browser (many of them have installed (and use) more than one; many people don't even realize they have installed (and use) more than one because browsers and their components are used behind the scenes in all kinds of shapes and forms.

c) it defines the problem space in terms of SIP reachability - then un-surprisingly concludes that we need to reach
devices via SIP.  

Most of those users are directly reachable via the GSM Um radio interface - but I don't see  any need to add Um constructs to rtcweb. Or looked at different way - the stated group perspective is wrong IMHO .

Tim (speaking for no-one but himself).