Re: [xsl] document() function and error-handling

Subject: Re: [xsl] document() function and error-handling
From: Rolf Ade <rolf@xxxxxxxxxxxx>
Date: Sat, 05 Jan 2008 04:39:00 +0100
Michael Kay wrote:
>> I'm really not trying to beat a nearly-dead horse here (I
>> swear!), but the XML spec says in '2.6:
>> [17] PITarget ::= Name - (('X' | 'x') ('M' | 'm') ('L' | 'l'))
>
> This rule says that the name XML in any combination of upper-or-lower case
> is forbidden. There's a rule elsewhere that says names starting with "xml"
> are reserved for future standardization.


That's getting off-topic and I already hear the barking. Though, and
sorry, that's a bit wish-wash. I read the XML rec as Scott (and
implemented that way) that the name XML in any combination of
upper-or-lower case is forbidden. Longer names starting with the chars
XML in any combination of upper-or-lower case are allowed by this
rule. I'd love to see a reference to the "rule elsewhere". There's a
rule, that every qname starting with a prefix xml isn't OK. But we're
talking about a name here, not a qname. Isn't XML about to have a
sharp sword to distinguish well-formed from not well-formed?

> Some parsers will give you a warning if you use such names;

Do they? And if in fact a few to name do, with what reason?

> and your application may break if W3C decides to attach a special
> meaning to the name you have chosen for yourself. Since there's an
> infinity of names that don't begin with "xml", it costs you nothing
> to avoid this risk.

Sure. Nobody knows, what the w3c decide. But anyway. If everything
else works, such a PI name isn't a problem now and the next future.

rolf

Current Thread