[
Lists Home |
Date Index |
Thread Index
]
On Wed, 5 Feb 2003 02:40:27 +1100, Rick Jelliffe <ricko@allette.com.au>
wrote:
>
> Databases servers manipulate data in some binary structure.
Uhh, XML database servers manipulate XML :-)
> that middleware should talk some optimized proprietary code to its
> (tightly-coupled) server, with the interface
> to the Database at the server being document objects (or whatever XML-ish
> thing is needed.)
Well, the problem I see (and I admit I see it on the horizon) is that XML's
value proposition is that it is general-purpose, vendor-neutral, platform-
neutral, language-neutral, etc. XML technologies are being used to
displace the "optimized proprietary code", because it is so much cheaper to
harness Moore's Law to get performance out of generic tools than to write
optimized proprietary code that actually works for the intended purpose.
So, XML is in some sense becoming a victim of its success The solution of
simply using XML as the interchange format between proprietary systems is
becoming less viable. XML is being thrown into the breach to replace
proprietary systems, and it demonstrably works well in the proof of
concepts. Now it's being deployed in the Big Time, and will be asked to
meet high performance and scalability requirements as well as
interoperability requirements.
> Which makes me wonder whether one of the value of binary formats is to
> provide a greater range of solutions to developers of tightly-coupled,
> probably-proprietary protocols?
The problem is that once you start drinking the open standards / vendor
neutral / interoperability nectar, it's hard to go back to the proprietary
stuff when the headaches start :-) If you've put XML everywhere as an
interop tool, it's both bad business and technologically short-sighted to
rip it out and put proprietary protocols in place. And if the customer is
happy with the proprietary stuff, what's the point of XML? It would be
*nice* if there were open standard(s) for things like XML subsets that are
easier to parse and/or alternative serializations of the infoset that are
more efficient to exchange over standard protocols. I suspect that a lot
of people are reinventing the same wheels, and it would be nice to see
standards in this area.
|
