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] [Fwd: Re: The 11-pound solution (fwd)]

[ Lists Home | Date Index | Thread Index ]

On Thursday 12 September 2002 21:18, Tim Bray wrote:

> [Snell]  If you have an XML-based infrastructure, then a device that can
> do XSLT-controlled XML to HTML transformations at network speed has a
> great deal of appeal to folks.  Again, I don't think the XDR or ASN.1
> analogy holds, although there's probably a niche market that would like
> a network device that did BER to DER.

That looks more like a followup to me, but hey, I'm not fussy.

> [Winer] No Dave, our first product is real, does work, and is solving
> customer problems.  You might want to take a look at the "customer
> successes" link on the same page that Tim posted. Oh heck, here's
> the link.  It's worthwhile because it gives more concrete information
> about why someone would buy and how they would deploy it.
> 	http://www.datapower.com/products/customer_successes.html
> As for Intel and their XML Director, shortly after killing the product
> they spent $500M buy iPivot, which they just spun as Tarari.  Perhaps
> they're just confused. :)

"Business Problem: Hemscott adopted XML to drive revenues as it enables the 
financial services firm to supply more information to more clients worldwide 
than is possible with proprietary technologies. But XML's nested structure 
and verbosity introduced serious bottlenecks that rendered revenue-generating 
applications useless. Large file transformations took as much as 20 seconds 
to complete when using the open source Xalan processor on general purpose 
hardware. As a zero-latency-enterprise, Hemscott found that it's performance 
and scalability requirements soon outstripped the capabilities of any 
software-based solution."

Now when I started moaning at people that XML was too verbose and slow to use 
for data they all just said "Bah, processing power and bandwidth are cheap, 
machines have more than enough CPU to cope, you're just like those guys who 
opposed high level languages taking over from assembler"... now companies are 
needing to shell out on specialist hardware to do things with XML? *rolls 
eyes* :-)

Now, as I said, my NFS server at home is an ancient Sun IPX (1989 or so 
vintage) that quite happily floods a 10Mbit interface with information it has 
encoded in XDR on the fly (particularly if I run sprayd!) without needing an 

So how about we stop trying to [ab]use XML for data and just bind an XSLT 
engine to a nice XER (http://asn1.elibel.tm.fr/xml/) codec, if you REALLY 
want XSLT (which I don't having had far too much trouble getting it to do 
what I want, but not everyone's trying to solve hard problems), hmm? :-)


A city is like a large, complex, rabbit
 - ARP


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

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