Re: [dispatch] HELIUM at Dispatch

[Ben Schwartz <[e-mail]>]

On Wed, Jul 11, 2018 at 3:15 PM Adam Roach <[e-mail]> wrote:

> [as an individual]
>
> This is small, but it's going to be a recurring papercut if this works
> gains traction: I'm having a really hard time reading a document talking
> about "HIP" that is in no way related to
> https://datatracker.ietf.org/wg/hip/about/ -- please consider renaming
> this protocol.
>

OK, will do.

It would probably also be worthwhile explaining how the features of HELIUM
> are significantly different than RFC 1928. I see raw IP tunneling and
> WebSocket support as differentiators, but it's not clear that this requires
> restarting the protocol from ground zero rather than extending an existing
> extensible protocol.
>
> Lest my suggestion above be perceived as tongue-in-cheek, I'm dead
> serious: RFC 1928 is used extensively by the Tor network, and there is a
> fair body of deployed code in modern use, at least in that context.
>

I've written a lot of code for RFC 1928 (SOCKS 5) myself, including the
rarely used SOCKS5-UDP support, so I fully appreciate this.  Alternatively
TURN also supports UDP, and with RFC 6062 can proxy TCP too.  I do feel
that a clean-slate design is worthwhile in this case, but you're right that
I haven't really tried to argue that in the draft.  I'll think about the
right way to make that claim.


>
> /a
>
> On 7/11/18 12:27 PM, Ben Schwartz wrote:
>
> Hi Dispatch,
>
> One of the drafts on the Dispatch agenda for next week is HELIUM:
> https://tools.ietf.org/html/draft-schwartz-httpbis-helium-00.
>
> There has been some discussion on the HTTP list (e.g.
> https://lists.w3.org/Archives/Public/ietf-http-wg/2018JulSep/0041.html),
> but the most detailed presentation on this proposal will be here in
> Dispatch.
>
> HELIUM has relevance to privacy protection, P2P through NAT, VPNs,
> advanced IP features, QUIC, and congestion control, among other topics.  I
> hope anyone who's interested will read the draft (or at least the
> introduction) before our session on Monday.
>
> Thanks,
> Ben
>
>
> _______________________________________________
> dispatch mailing [e-mail]://www.ietf.org/mailman/listinfo/dispatch
>
>
>