RE: Namespaces 1.1 Last Call -- I18N WG comments

[I'm copying www-tag, so that the issues are widely aired]

Hello Richard,

> -----Original Message-----
> From: Richard Tobin [] 
> Sent: 11 November 2002 15:48
> To: Misha Wolf
> Cc:;
> Subject: Re: Namespaces 1.1 Last Call -- I18N WG comments
> I was just looking at your proposed IRI identity examples, and I see
> that they all have an ampersand in the middle (both red&rose and
> CV&resume).  These would have to be & or similar in the actual
> attributes, which seems to complicate the examples unnecessarily and
> obscure the main differences.  Did you really mean to use ampersands?

We discussed your mail at our telcon yesterday and our response is ...

Firstly, please do *not* rely on the mail archive for our examples, as
the archive doesn't correctly display the text of our message.  The 
archive link is:

The text of our messages as originally sent is:

| Dear xml-names-editor,
| On 28/09/2002 13:53:10 Misha Wolf wrote:
| > On 27/09/2002 21:04:18 Misha Wolf wrote:
| > > Dear xml-names-editor,
| > > 
| > > This mail contains some of the comments of the I18N WG on:
| > >    Namespaces in XML 1.1
| > >    W3C Working Draft 5 September 2002
| > 
| > [The suggested examples below have been extended to include a decimal|
  > numeric character reference, and have been slightly re-arranged.]
| > 
| > Suggested examples:
| [...] 
| It turns out [blush] that we had spelt "resumé" incorrectly.  It 
| actually has two accents and so isn't ideal as an example.  So we 
| propose "red&rosé" rather than "CV&resumé".
| Suggested examples:
|    Consider:
|       xmlns="é"
|       xmlns="é"
|       xmlns="é"
|       xmlns="é"
|       xmlns="é"
|       xmlns=""
|       xmlns=""
|       xmlns=""
|       xmlns=""
|       xmlns="É"
|    where "é" has been defined to be "é".
|    The first 5 are identical for the purpose of namespace reference 
|    matching and have the normalized attribute value [1]:
|    The remaining 5 differ from the first 5, as well as from each other,|
     and have the following normalized attribute values, respectively:
| [1]
| Misha Wolf
| I18N WG Chair

We stand by the issues highlighted by our examples.  If the XML Core WG
wants to highlight (all of) these issues using a larger number of 
simpler examples, that's OK with us.

The motivation for the high level of detail is that the original
Namespaces spec doesn't define its terms (eg "character" or "exactly the
same character-for-character") and doesn't provide examples to
illustrate the many gotchas.  We mustn't make the same mistake this time

The issues we have sought to highlight are:

1.  The value of a namespace name is obtained by applying the steps 
    described in:
       3.3.3 Attribute-Value Normalization

2.  Identity between namespace names is determined by doing a binary 
    match on the results of those steps.

3.  Consequently, the following *are identical* (where "é" is a 
    reference to an entity containing "é"):
    and the following *differ* from the above and from one another:

4.  Furthermore, given:
    the corresponding namespace name is:é

for the I18N WG

> -- Richard

------------------------------------------------------------- ---
        Visit our Internet site at

Get closer to the financial markets with Reuters Messaging - for more
information and to register, visit

Any views expressed in this message are those of  the  individual
sender,  except  where  the sender specifically states them to be
the views of Reuters Ltd.

Received on Wednesday, 13 November 2002 11:19:46 UTC