Re: DataChannels and Streams: Anything for TPAC?

As I said in Stockholm, I think allowing data channels in service workers opens up some huge
opportunities. 

Specifically I’m interested in the ability to use the two to facilitate web-services from behind NAT.

Combining the data channel’s NAT traversal with service worker’s ability to present pages to the
browser as if they had been fetched over http(s) allows devices behind NAT to act as web servers.

My feeling is that this presents a major opportunity for a more decentralised web.

I’d be happy to present this if there is interest (and a guest invite).

Tim.

 

> On 4 Oct 2018, at 18:29, Peter Thatcher <pthatcher@google.com> wrote:
> 
> While there is a discussion around and an issue for WHATWG streams for the RTCQuicTransport (https://github.com/w3c/webrtc-quic/issues/2 <https://github.com/w3c/webrtc-quic/issues/2>), is there really that much interest for WHATWG streams for SCTP data channels?  
> 
> I recall there being much more interest for service workers and SCTP data channels.  If we're going to talk about SCTP data channels, it seems like time would be better spent talking about service workers rather than WHATWG streams.
> 
> That said, I don't know if anyone has updates for either topic.
> 
> 
> 
> On Thu, Oct 4, 2018 at 8:22 AM Harald Alvestrand <harald@alvestrand.no <mailto:harald@alvestrand.no>> wrote:
> In setting up the agenda for TPAC, and looking at what we talked about
> in Stockholm, we're missing one big item:
> 
> In Stockholm, we had people express a strong desire to fit a
> Streams-based interface onto datachannels (both Stream-ing a single
> message and Stream-ing a datachannel; the latter might end up as a
> Stream of Streams....)
> 
> 
> We haven't seen any followup on this since then, so our current thought
> is that this won't be a TPAC agenda item.
> 
> What do people think here?
> 
> Harald, for the chairs
> 
> -- 
> Surveillance is pervasive. Go Dark.
> 
>