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] Re: Where does the "nothing left but toolkits" mythcome fr

[ Lists Home | Date Index | Thread Index ]

Bob Foster wrote:

> So if one called it BML and wrote SAX and DOM APIs to read and write it, 
> you'd be happy with it?

That's still a little too close to comfort. For one thing, such a binary 
format is not markup and probably not a language.

Come up with a completely different name, such as FIF (Flexible 
Information Format) that does not remind anyone of XML; and don't 
reference XML normatively or not in the core specs. Let it succeed or 
fail on its own merits. Then I'm happy.

If someone wants to write DOM or SAX adapters for this new format, fine. 
Certainly people have written SAX adapters for all sorts of weird things 
such as SQL databases. There are problems when they do that. For 
instance a lot of apps that consume data from SAX implicitly assume 
things such as element names don't contain white space, and can get 
tripped up when SAX adapters don't enforce these conventions. But I can 
live with that.

Still, I think it might be more productive for the FIF community to 
define their own APIs that better fit their data model than the XML data 

Yet another problem with inviting the binary community under the XML 
tent, is that they're not going to be satisfied with an alternate 
encoding for XML. This is only the tip of the iceberg. As soon as the 
binary folks get their spec approved, they're immediately going to want 
to change SAX and DOM and XSLT and everything else with just a few more 
extensions so they can pass around the binary data in its native form. 
After all, why pay the cost of converting to and from text all the time? 
(We already see this in XOP.) Then they're going to say that all the 
overhead of Unicode and BOMs and the like is just killing their 
performance, and can't we just cut out these pieces for the legacy 
parsers? Before we know it, plain vanilla XML is a distant but pleasant 
memory as we all struggle to deal with the mass of messy APIs to control 
the opaque binary data passing through our networks.

Elliotte Rusty Harold  elharo@metalab.unc.edu
XML in a Nutshell 3rd Edition Just Published!


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

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