[UA-discuss] Technical Constraints to UA & EIA success

[Don Hollander]

Andrew,

We have no issues with tools that will block some (or even all) domain names and email addresses.  We just want to make sure that their blocking criteria is conscious and not accidental because their software rules made assumptions about the nature of all domain names and email addresses.

So, you don't think the issues around shortcomings in IDNA 2008 are technical?

And you don't know of any outstanding technical issues pending resolution?

D



-----Original Message-----
From: UA-discuss <ua-discuss-bounces at icann.org> On Behalf Of Andrew Sullivan
Sent: Wednesday, 27 June 2018 12:56 AM
To: ua-discuss at icann.org
Subject: Re: [UA-discuss] Technical Constraints to UA & EIA success

On Tue, Jun 26, 2018 at 07:17:51AM +0000, Don Hollander wrote:
> There are some issues around IDNA and Unicode versions.   (How far did PRECIS
> go do addressing these?)

PRECIS did nothing about them, because it wasn't chartered to work on IDNA.

The problem there is really a gap between UTC's approach to code points and what IDNABIS thought it was getting from Unicode.  This appears to be an unbridgeable gap, but it's not really a technical
problem: it's a political problem of different SDOs working in the same space.  The cases that are at issue, however, are so abstruse as not to be serious effective barriers right now anyway.

> So, what do you think are the remaining technical constraints to 
> acceptance of all domain names and all email addresses?

I think the goal is unreasonable, because _all_ domain names and email addresses have never been accepted anywhere.  Blacklists and blocklists have been the reality on the Internet since before ICANN even existed, because people and enterprises have always wanted to get their work done without connecting to everything.  But if you mean _general_ acceptance of all domain names and all email addresses, the piece still missing are effective policy engines that can be used in clients, so that they can handle names that the user can't really use (i.e. domains in a writing system the user can't read or otherwise parse).  This is, in some sense, the difference between an _identifier_ and a usable name, and it's a vexing problem.  Every culture has a story like the bliblical Babel narrative, and it's not for nothing!  This may not be a technology we can invent.

A

--
Andrew Sullivan
ajs at anvilwalrusden.com