Re: Multiple substrates (Was: Re: sasl in beep or payload)
"David Chizmadia" <[email protected]> Wed, 19 Dec 2001 13:11:57 -0500
| Newsgroups | gmane.ietf.sacred |
|---|---|
| Message-ID | <005801c188b8$a6115bc0$a000a8c0@chizmadia> |
Stephen,
> So it seems that what's been suggested so far is along the lines of:
>
> 1) only define beep
> 2) only define http
> 3) define both in one rfc, client picks one, server MUST/SHOULD/MAY do
> both (if not MUST do both, then one of beep or http has to be
> mandatory-to-implement)
> 4) define both in one rfc, client and server pick what they like (yuk!)
> 5) first do 1) but then do 2) in a separate rfc, re-using as much as
possible,
> and making it as easy as possible for a single server to implement
> both
>
> Are there other combinations/substrates/ways-to-organise this that people
> would like to suggest before we get into the repsective (de-)merits of
> these?
6) in a distinct RFC, define the normative SACred
PDU content and format, as well as a set of
mandatory requirements for the substrate that
carries those PDUs. Define additional RFCs for
beep and http. Let other SDUs (like OMG) reuse
the PDU RFC with our substrate (GIOP/IIOP).
> Stephen.
-DMC