recommending URIs, A-labels and LDH-labels in HTML

[John C Klensin]

--On Thursday, 31 January, 2008 16:23 -0800 Erik van der Poel
<erikv at google.com> wrote:

> On Jan 31, 2008 12:53 PM, John C Klensin <klensin at jck.com>
> wrote:
>> (2) You should push, where possible, for URIs and IRIs in the
>> documents you are indexing to contain final-form names
>> (A-labels or U-labels).  There really should be very little
>...

> Yes, we could do something like that in our Webmaster Help
> Center. Actually, one reason we would give for using URIs and
> A-labels (and LDH-labels) is that IE 6 does not support IDNA
> and there are still a lot of IE 6 users.

There is one extra issue with this that may go with the advice.
For years, it has been a common practice for HTML pages to
contain links constructed approximately like

  <a href="http://foo.bar.baz/">http://foo.bar.baz/

I keep expecting browsers to start testing for identity of those
two strings as an anti-phishing mechanism and popping up
warnings when they appear to be slightly different (slightly
different is the problem, lots different is not, note

  <a href="http://foo.bar.baz/">[Click here]

).  We had better be sure that everyone understands that

  <a href="http://A-label1.A-label2.A-label3/">
     http://U-label1.U-label2.U-label3/

should not trigger nasty warnings if the domain names are
label-for-label equivalent.

     john