[idn] Re: I don't want to be facing 8-bit bugs in 2013

[Paul Robinson <paul@iconoplex.co.uk>]

On Mar 20, "D. J. Bernstein" <djb@cr.yp.to> wrote:

> False. IDNA does _not_ work. IDNA causes interoperability failures. Mail

>... with the current DNS resolvers in place...

OK, others have pointed out failures with things like SSL/HTTPS (which is
broken in several interesting way anyway from the point of view of
scalability and long-term usefullness), but even then I can see work
arounds. All of them are more preferable to throwing away every serial
console in the world because a greek delta doesn't display properly on it
and therefore has a 'display failure'.

> That assumption is false. Consider, for example, an MTA configured to
> accept mail for pi.cr.yp.to, with a Greek pi. The MTA compares the
> incoming domain name to pi.cr.yp.to. That doesn't involve the resolver.

Well, funnily enough I have to tell my MTA about .co.uk domains AND .com
domains as well, even though the bit in front of them is IDENTICAL. If I
register the .net I have to tell it about that as well. They're different
domains. If I buy an IDN with a Greek letter in it, I'm going to have to
tell my MTA about that. If think that by adding a new domain to your config
you shouldn't have to tell your MTA about that, you're obviously not a very
experienced admin.

> Now, please explain why the same user should prefer a domain name that's
> _occasionally_ displayed with the desired delta but _usually_ displayed
> as incomprehensible gobbledygook.

Until he fixes his machine to have proper Greek alphabet support. However,
when it comes over to my PDP-8 it appears as gobbledygook. But it still
works. Sure, that's REAL broken.

> In short, you're looking at the long-term IDNA benefits (never mind the
> interoperability failures and all the other problems) but refusing to
> look at the long-term UTF-8 benefits. Inconsistent once again.

The IETF is about the long-term. It not about the next 2 years. Even I know
that. You should certainly know that.

>    * interoperability failures;

That don't exist. Or the ones that do can be given work-around far simpler
than replacing every piece of equipment in th eknown universe and updating
every piece of software ever written. Ever.

>    * inconsistent displays of the same name;

That don't matter.

>    * unnecessary implementation and deployment costs;

The fact you're suggesting that what *I'm* proposing and what the IDNA is
proposing requires unnecessary implementation and deployment costs, quite
frankly, makes me snort my coffee up my nose.

>    * multiple semantically similar names;

OK, let's not allow uppercase Greek Alpha. Does that fix one of your
problems? Sheesh.

>    * identical displays of different names; and

Not a problem for the IETF. Not even a problem for users. In Greece A means
Alpha. In the US it means 'Ay'. If you're a Greek in New York, you might
have problems. You're going to have problems anyway, but at least with IDNA
PunyCode is your friend.

>    * typing failures.

We get those anyway. It's why slahdot.org exists.

> False. Every step in http://cr.yp.to/proto/idnc3.html preserves
> interoperability.

Fixing half a dozen pieces of software and then just allowing every mand and
his dog to register special Bernsteinised-domains? Right. I see.

> There are several options. One option is to work around the hardware
> limitations in software, displaying something like
>
>           |
>    /\/ /\ |   /^ /\ /\ /\
>    \/\ \/ | * \_ \/ | | |

*ROFL*

# cat Mail/inbox | utf8decode | figlet

That's how I like to read my mail in the morning!

Seriosuly man, I'm sat here on a customer's site in the middle of the South
Atlantic looking forward to an 18 hour RAF flight home to the UK, and I'm
crying with laughter. You really have cheered me up. That was so
off-the-wall and unexpected - to propose IDNs to be displayed as ASCII art -
that I'm going to be smiling and giggling to myself for days. You have
confirmed what I should have known - you're stark raving insane. Quite,
quite, quite mad. Fantastic.

> Another, much more popular, option is to move your email reading, web
> browsing, etc. from your 1970s-vintage VT100 to a graphics terminal.
> Have you considered the VT340, for example? Or an IBM PC, model 5150?

You're proposing the abondonment of legacy hardware when there is no need.
That costs money. Serious money. If I have my hostname setup as
pi.alpha-ol.com and I need to gain access over the serial port at boot with
a VT100, the option of anything else is not available.

> > a good 20% of sites out there will just have to shut down ops
permanently
>
> Get a grip, Paul.

I'm not the one proposing we turn the world into one great big ASCII art
factory.

Anyway, I'm going to be away from mail for a few days due to travelling
around - don't think I'm giving up on you yet though. :-)

Seriously... ascii art... hehehe...

--
Paul Robinson