Re: [xml-dev] Ten Years Later - XML 1.0 Fifth Edition?

[Jonathan Robie <jonathan.robie@redhat.com>]

Elliotte Harold wrote:
> 1. Please do remember we're talking only about name characters, not 
> the text content of a document. The XML 1.1 proponents kept confusing 
> this issue the last time around too.

I'm not an XML 1.1 proponent, in fact, I would love to see XML 1.05 kill 
XML 1.1, but I do feel that strings that represent words in human 
languages should be legal names. Regardless, that decision was made 10 
years ago for most languages, we're talking about whether it also 
applies to languages for which Unicode has added support in the last 10 
years.

> 2. These characters are only a good idea if the documents are only 
> exchanged among speakers who all share the alphabet/syllabary/etc. and 
> the necessary fonts.

Would XML have been a better language if XML element and attribute names 
were limited to English, and non-English character sets could be used 
only in content? That's probably where this principle would lead.

And my guess is that the answer depends somewhat on whether you are 
American, Japanese, or Amharic. And again, this decision was made 10 
years ago, it's too late now.

> 3. The real issue is undefined Unicode characters that this proposal 
> allows, not well-defined characters at all.

Can you say more? I thought XML 1.1 was a real mess on this front, but 
XML 1.05 is more careful about the characters it allows.

> 4. The other issue is well-defined characters that are not wise for 
> use in XML names, such as easily confusable characters, and various 
> punctuation marks and spacing characters.
>
>> I'm guessing you didn't mean that to say what I just understood it to 
>> say. Surely you don't object to me putting Greek in a file even if 
>> you personally can't read it.
>
> Actually, I personally can read it. :-)

Cool!

Jonathan