[FoRK] binary XML

Eugen Leitl eugen at leitl.org
Sat Jan 22 00:52:55 PST 2005

On Sat, Jan 22, 2005 at 12:45:42AM -0500, Gavin Thomas Nicol wrote:

> >Interoperability? If your partner is using XML already, a binary 
> >drop-in
> >replacement which is far more infrastructure replacement is an easy 
> >sell.
> This is purely based on network effects... the technology as such 
> doesn't
> buy you anything, and indeed, may cost you.

I'm not sure whether you're talking binary XML or vanilla XML.

Whenever you've got modular systems and need well-debugged code on the cheap
(free) instead of reinventing a polygonal wheel by braindead coding monkeys, 
XML shines. This doesn't require network effects beyond a single company.
Making servers talk XML to each other or modules dumping stuff to be parsed
by other modules via file system is already very effective, and frequently
warrants a rewrite of old cruft.

Network effects appear if somebody on the other end wants to e.g. SOAP with

If you've got a binary-XML painless (a la apt-get install or CPAN shell) 
drop-in replacement you can benefit by reducing the amount of iron in the
rack (or not populating the rack that much in the first place). Such
migration is easy to test and to deploy.

If people on the other end do the same (it's an easy sell, if you can get an
order of magnitude improvement in throughput with the same equipment, or
scale far better). All assuming, of course, full 1:1 XML compatibility with
open source tools. 
> I have no problem selling XML technology so long as people don't 
> confuse network effects
> and technological applicability: that is where the benefits of XML 
> really shine. A lot of
> real hype doesn't acknowledge that though... the hype makes XML itself 
> into a magic technology,
> which it is not.

No disagreement here. No pointy-hairs present, after all.

Eugen* Leitl <a href="http://leitl.org">leitl</a>
ICBM: 48.07078, 11.61144            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
http://moleculardevices.org         http://nanomachines.net

More information about the FoRK mailing list