IEA BoF

[Leslie Daigle <leslie@thinkingcat.com>]

This was the "Internet Email Address" BoF.  

There were a couple of proposal drafts going into the BoF session,
but the bulk of the discussion (appropriately) focused on the
larger question of which was the right approach (not the specifics
of the proposal).

The proposal directions were:

	. do an IDNA-like 8-bit to 7-bit conversion hack (Paul Hoffman)

	. give up on preserving the existing syntax remnants, and give
	  people what they really want, in the shape of message target
	  addresses from an unrestricted character base  (put forward
	  as a question in John Klensin's proposal)

	. just send (UTF)-8

The BoF chairs (Patrik Faltstrom & Pete Resnick) made it clear that
the desired outcome of the BoF was not necessarily a WG proposal,
that there might be work more appropriate for the IRTF.

The sense of the room seemed to be that the unfettered internationalized
identifier was too big a problem space to hope to solve (in an engineering
or a research group), that there is desire to have internationalizable
message addresses on business cards sooner rather than later, and
yet that doing the IDNA-like (or some other compatibility mapping)
might have hidden perils and not really solve the desired problem.

The meeting ended without conclusion about where to go from here, with
the session chairs calling for more discussion on the mailing list.

I'll say here the same thing I said in the room -- deciding if or
how to proceed here has to take into account that the "cheapest"
route (IDNA-like compatibility mapping and coping) path might
have *significant* collateral costs.  For example, we haven't yet
understood entirely how to get IDNs and domain names to play
together in expected ways in protocols (i.e., that document that
Ted started and I'm supposed to finish for the IAB...); what
would be the similar effects here?  What does this do (or not do)
for other user identifier systems that have been modeled on e-mail
addresses, such as SIP addresses?  Would proposals that were cheap
for e-mail be equally successful elsewhere?  etc.

So, yes, doing this properly might be very difficult and expensive,
and yet perhaps not *the* most expensive.

Leslie.


--

-------------------------------------------------------------------
"Reality:
     Yours to discover."
                                -- ThinkingCat 

Leslie Daigle
leslie@thinkingcat.com
-------------------------------------------------------------------