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

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   RE: [xml-dev] Some Questions on XML and XML Processing

[ Lists Home | Date Index | Thread Index ]


From: Ram Menon [mailto:menon_dev@hotmail.com]

>1) Is the way XML is written[strctured, tree fashion] just because of the 
>structured nature of the data it represents? Why was this mode of 
>representing a document chosen in first place?

Because XML is the most recent of a series of developments starting 
in the late 1960s with GenCoding to represent document structures in 
a way that a computer could process them without intervention and 
a human could construct and edit them.  Documents are traditionally 
hierarchical.

>2) XML is so verbose that it cannot be easily interpreted by a human 
>audience.

That is the same mistake Tim Berners-Lee made.  Markup systems were 
designed precisely so a human could interpret them.  This works well 
in fact but works better or worse depending on the design of a given 
document type and the familiarity of a given human with the names and 
structures provided by the document type designer.

>Why is it that there aren't two versions of an XML document - a "direct" 
>human readable representation of the XML content [ other than using XSLT
and 
>making it readable], and another representation for processing it, which is

>compact and available for fast processing ?

Because lots of experience has shown that optimization pays little back 
in performance and raises the cost of management.  The point of view of 
performance has to be balanced against other dimensions of production.

>3) Why is it that parsing an XML file leads to only a single XML Document?
>There could be a message coming in, which might mean to represent multiple 
>payloads; i.e.lets say three XML Documents.... the xml processor could 
>interpret these, and after parsing the document, spit out three parsed XML 
>documents?

SGML originally had features for subdocs and there was work going on in 
the mid-90s on packaging documents. Your scenario will run into the problem 
of knowing that a single package has three documents, how to pack and unpack

these, and so on.  This is still an area in need of firm practice and 
standardization.  Still, applying transforms, nothing prevents doing 
what you suggest other than configuring the local system to produce 
three documents from the single one you provide as a Message.

In short, no two people agree much about what a 'document' is except 
that XML is a document model.

len




 

News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS