[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
Re: [xml-dev] json v. xml
- From: Rick Jelliffe <rjelliffe@allette.com.au>
- To: xml-dev@lists.xml.org
- Date: Fri, 05 Jan 2007 16:48:44 +1100
Michael Champion wrote:
> There's also the "zero-one-infinity rule".
Which, if applied to life, is a religious statement not a game-theoretic
statement as it may appear.
If I go to a supermarket and tell the startled shoppers to only buy no
soup, one flavour only, or every flavour only, because I subscribe to
the "zero-one-infinity rule", I will probably be ejected. "They can have
any colour they like as long as it is black" came because of the drying
properties of the new paint of the time which happened to be black IIRC,
not because of the zero-one-infinity rule.
> There is something to be said for trying to unify all this around some sort
> of "XML 2.0" that is founded on a common, minimal but extensible tree data
> model rather than the bits on the wire [zipping up flameproof suit in case
>
Well, that would be SGML 2.0 wouldn't it? SGML was founded on the
notion that
people had legitimate needs for different syntaxes, and that what was
important was
being able to describe the syntax differences in a way that would allow
a common element structure information set to be extracted. (XSD uses
grammars because XML DTDs use grammars because SGML DTDs use grammars
because SGML actually used them for parsing, not because grammars are
particularly adept at modeling data.)
A modern SGML 2.0 would allow JSON, XML, Wiki, ASN.1 and LDAP to
be reconciled with a common API. Eric van der Vlist has been doing API
work along
these lines in recent times. (The SGML approach was, however, that
instead of doing things in terms of a common tree model, you do things
by syntax, of course.) However,
an advantage of JSON is that don't need any special API at all: a nice
quality
that S-expressions have and, I suppose, the XML Beans approach also may
have for
Java.
> Just perhaps, the threat of
> JSON/EXI-induced chaos might motivate the W3C to think about this.
Why think in terms of chaos? There is no chaos that some people use GIF
and some people use JPEG. A rich solution space is good. The chaos comes
from when the individual technologies are so bloated and indivisible
that people cannot compose their own solutions using optimal
subtechnologies.
Three nice fat ISO standards are SGML (ISO/IEC 8879:1986), ASN.1
(ISO/IEC 8824:1987 and more) and ECMA Script (*ISO*/IEC 16262:2002).
People who use them have profiled them for use on the web: XML, FastXML
and JSON. The profiles from actual use, once established, get reflected
back into the formal standard (WebSGML, XER, ???) in time. I don't see
it as chaotic, its healthy and inevitable.
Cheers
Rick Jelliffe
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]