Type and Structure Re: ASN.1 and XML

[Rick Jelliffe <ricko@allette.com.au>]

From: "Jay Zhang" <jayzhangsj@hotmail.com>

> Hi, Rick,
> 
> You made a good point about "informality"
> heritage of XML. A curious student using notepad
> can go a long way with XML, but not much with
> ASN.1. HTML has prepared large number of people
> for XML.
> 
> The point I like to contribute is the subtle
> difference between type and structure. XML Schema
> and ASN.1 have the same problem: too strong a type
> flavor. XML allows a wide spectrum between a simple
> structural checking to a pseudo-typing system while
> ASN.1 is a rigid typing system.
>
> Well-formedness is pure structural. Schematron is
> mostly structural flavor with isolated validity
> checking. RELAX, TREX, ... are more of type systems.
> XML Schema seems to be achieving a pure typing
> system.

Yes. The grammar-based schema languages are trying
to interpose some type system between markup structure
and the model, without any guarantee that their type systems
will be any easier to map to a model than the original structures
were. Witness XML Schemas, which is conceived in terms
of "components" (its model) but has no analog of "component"
in either its type system or its markup. TREX and RELAX 
seem to snuffling for diamonds in the same sandpit, with as
much chance for success.

> It has always been heartening every time when I
> see the comparison of XML and ASN.1.
> 
> With Rich Li, I wrote one of the most widely used
> ASN.1 tool: the ASN.C/C++ from DSET, before I left
> in 1991. When I read about XML back in 1997, the
> first thing I did what writing an e-mail to Dan
> Shia, then Chairman of DSET, that everything we
> liked to do with ASN.1 was to be done in XML.

Wow, an ASN.1 pioneer in our midsts.  Excellent.

> DSET was not set up to be so narrowly focused on
> Telecom tools later or even more narrowly, and
> disastrously, on CLEC now.
> 
> When OMG was working on the CORBA standard, DSET
> made a proposal that loosely contains SOAP based
> on ASN.1. We got two votes, one from Microsoft
> casting vote for every proposal and another from
> ourselves. The rest is history.
> 
> Conceptually, the idea of robust data encoding is
> the same in XML and ASN.1. That is why there is so
> much passion in ASN.1 people about XML.
> 
> However, I personally do not think ASN.1 as it is
> can be a real competition for XML, unless XML is
> completely absorbed into ASN.1.

That is interesting. Do you think pivoting through ASN.1
would be a good way to get optimised encodings for tree
data?  Or is ASN.1 still to far from the conceptual model
to represent any actual advance: is it just a "lateral arabesque"
(a decisive advance sideways.)
 
> Please repost to xml-dev if you see fit. I have
> problem posting.
 
Done
Rick Jelliffe