Re: Fwd: New Version Notification for draft-wouters-edns-tcp-chain-query-00.txt
Joe Abley <[email protected]>
| Newsgroups | gmane.ietf.dnsext |
|---|---|
| Message-ID | <[email protected]> |
On 2013-09-11, at 11:56, Paul Wouters <[email protected]> wrote: > On Wed, 11 Sep 2013, Joe Abley wrote: > >> So, perhaps an EDNS0 option that allows some degree of capabilities negotiation (client says I can do { x, y, z }, server may provide responses with similar options, client issues test query over first message sent on newly-opened TCP socket to catalogue capabilities). >> >> This might also be a way to signal the availability of other transports like SCTP or HTTP. > > I'd prefer an option for just TCP so this doesn't end up getting > bikeshedded. The client is (presumbly) already using TCP to connect. I'd prefer a more general option that would allow more connection parameters to be advertised in future, if/when they become interesting. Assigning one code point to "persistent TCP connection capable" could leave the door open in future for bikeshedding, but since that seems inevitable anyway, what's the harm? I'm keen/able to contribute text to such a document (separate from edns-tcp-chain-query), if you'd like a co-author. Joe _______________________________________________ dnsext mailing list [email protected] https://www.ietf.org/mailman/listinfo/dnsext