[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: all group composition

• To: xml-dev@l...
• Subject: Re: all group composition
• From: Carlos Pita <carlosjosepita@y...>
• Date: Fri, 1 Apr 2005 20:12:42 -0300
• User-agent: KMail/1.7.1

In my case, the user will enter a subset from
a considerable number of properties and
order is not significative. So although
the properties are simple and intuitive per se
and the user has no no need to specify
or even know the entire set, there is
no natural ordering between them, so
imposing an ordering seems undesirable
to me and forces the user to put an eye
on the xml editor and the other in
a detailed element order specification.
A lot has been written in favor
of more declarative lenguages that do not
or try to not impose constraints unrelated
to the problem domain. Also, in this case
it will be harder neither for users nor
for programs to read/validate the "unordered"
properties cause no one has any kind of
semantic dependence on the value of any other.
They can all be understood / processed
independently. Well, anyway it's not SO
important. Perhaps I would switch to relax
ng.

Regards,
Carlos


> > It's actually not a favor to users to allow unconstrained order where 
> > order doesn't matter, e.g., in 'data records'. 
> 
> I think that's a matter of opinion and it's an area where the document
> designer should not be constrained by what the tools allow and don't allow.
> 
> My own recent experience of creating documents whose schema imposed an order
> that I found unnatural and impossible to remember was a very negative one. I
> actually ended up editing without a schema and writing a stylesheet to put
> the fields into the required order.
> 
> Michael Kay
> http://www.saxonica.com/
> 
> 
> 

• Follow-Ups:
  • Re: all group composition
    • From: Jeff Greif <jgreif@a...>

• Prev by Date: Re: XQuery Adoption Survey Results
• Next by Date: Re: all group composition
• Previous by thread: Re: all group composition
• Next by thread: Re: all group composition
• Index(es):
  • Date
  • Thread