[
Lists Home |
Date Index |
Thread Index
]
Alessandro Triglia wrote:
> Robin Berjon wrote:
>>Well, no, it's not. It it were, you'd be able to get two
>>people to agree
>>on the answer to the "of what?" question. XML is just a syntax.
>>Maybe
>>your application layer uses XML to encode some information,
>
> In the case of ASN.1, it is not the application layer that uses XML to
> encode information. It is the standard XML encoding rules (EXTENDED-XER)
> that use XML to encode the instances of abstract types.
Yes, but from XML's point of view whatever is sitting above it be it
Joe's XML Interface, an ASN.1 toolset, or some application's ad hoc
model, is an "application". That application may be something that makes
things simpler/more flexible/more abstract/whatever to another higher
layer that it considers to be the application, it's still an application
to XML. Each layer has its point of view, and that of ASN.1/XER is
naturally different from that of XML.
The reason I'm being boringly pedant about this is that in the binary
interchange debate it will be of crucial importance to figure out
whether one is better off with a binary syntax mapping to XML 1.1 (with
no data model, or however many data models one may want) or if what is
best is an encoding of some data model (Infoset, PSVI, QDM, etc), where
values of "better off" and "best" are heavily multidimensional. Given
how much room there is for disagreement, getting our terminology
straight early and often is likely to be helpful ;)
--
Robin Berjon
|
