OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [xml-dev] Re: determining ID-ness in XML



At 10:13 AM +1100 11/11/01, Marcus Carr wrote:

>Elliotte, this simply isn't reality. Try telling a bank using XML as a feed to
>their processing back-end that they need to add a new attribute to 
>every element
>in their DTD - the IT department will laugh you out of the building. Don't
>bother trying to tell them that it won't break the existing system - 
>they won't
>take the chance. For organisations in that situation (and I know 
>firsthand of a
>number of them), the proposed solution of attributes is not appropriate. They
>won't do it, especially when you can't provide them with one single benefit to
>them. (Helping third parties isn't reason enough.)
>

OK, one more time since people keep failing to hear this. I would 
never tell a bank that. If they don't want to change their DTDs, 
that's perfectly OK with me. They can stick with their existing 
documents and DTDs if these meet their needs. Nobody has to change if 
they don't want to. In this case, they don't get to use 
xml:id/xlink:label in their valid documents. That's fine. We are 
*not* trying to come up with a solution that works for those 
unwilling to make any change to what they're doing now. In fact, such 
a solution is impossible.

You should also realize that whether we pick a processing 
instruction, an internal DTD subset, or just about anything else, 
none of this will work for organizations that are not willing to 
update their software and systems to take advantage of the new 
semantics of whatever is eventually defined, and that's OK too. 
Processing instructions may solve the very minor issue of invalid 
documents in organizations without the abvility to upgrade a DTD. 
However, they do nothing to solve the much bigger problem of 
providing software that understands the semantics of 
xml:id/<?xml-idatts?>.

All that's proposed here is defining an XML application to allow 
elements to be named so we can address them by name. Like XLink, this 
would probably be an attribute based syntax that layers cleanly on 
top of existing XML applications. In fact, depending on what's 
decided this naming scheme might even be XLink. And like XLink, it's 
totally optional. If it seems useful to you, use it. Otherwise ignore 
it. However, since there are a lot of people who would like to do 
this, it seems useful to have a standard means of providing this 
functionality rather than reinventing it in a dozen mutually 
incompatible ways.
-- 

+-----------------------+------------------------+-------------------+
| Elliotte Rusty Harold | elharo@metalab.unc.edu | Writer/Programmer |
+-----------------------+------------------------+-------------------+
|          The XML Bible, 2nd Edition (Hungry Minds, 2001)           |
|              http://www.ibiblio.org/xml/books/bible2/              |
|   http://www.amazon.com/exec/obidos/ISBN=0764547607/cafeaulaitA/   |
+----------------------------------+---------------------------------+
|  Read Cafe au Lait for Java News:  http://www.cafeaulait.org/      |
|  Read Cafe con Leche for XML News: http://www.ibiblio.org/xml/     |
+----------------------------------+---------------------------------+