[Xml-bin] :-(

[Stephen D. Williams]

"KINNER,JASON (HP-NewJersey,ex2)" wrote:
> 
> Excuse me for being a little flippant, but I believe this falls in the
> category of "Not My Job".  In theory, a well-designed binary transport
> mechanism for XML would transparently plug in to a text-based processing
> application (via the standards APIs, including, but not limited to, SAX,
> DOM, JDOM, JAXP).  The fact that the storage or transport format is binary
> should be irrelevant to the application developer.  The process I see is:
> 
> 1) Develop using existing tools (text-based) to validate application
> 2) Think "Wow, I wonder if the binary implementation would make this
> faster?"
> 3) Test assertion in #2
> 4) Decide which implementation you will use in your specific application.

This is correct except for improvements that depend on an extended API. 
DOM, for instance, uses the paradigm that you create a node and then
plug it into the tree.  An obvious enhancement when you are avoiding
copying, memory allocation, etc. is to build nodes in place which
requires the API be rearranged slightly.


> This whole process will be facilitated by the MOM vendors including this as
> an XML transport option (okay, that's kind of presumptuous, but why not?).

This is a possible avenue that makes sense.  I am however attempting to
produce a good general solution.

sdw

> -Jason
> 
> -----Original Message-----
> From: Leigh Dodds [mailto:ldodds@ingenta.com]
> Sent: Friday, April 20, 2001 6:49 AM
> To: Al Snell; xml-bin@warhead.org.uk
> Cc: xml-dev@lists.xml.org
> Subject: RE: [Xml-bin] :-(
> 
> Its just as important to identify where binXML *isn't* of any use, so
> that an informed decision can be made about where it can be usefully
> applied.
> 
> _______________________________________________
> xml-bin mailing list
> xml-bin@warhead.org.uk
> http://lists.warhead.org.uk/mailman/listinfo/xml-bin

-- 
sdw@lig.net  http://sdw.st
Stephen D. Williams
43392 Wayside Cir,Ashburn,VA 20147-4622 703-724-0118W 703-995-0407Fax 
Dec2000