Re: ASN.1 and XML

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

Does being pro-XML thereby make one anti-ASN.1?

What ideas do they have in common:
 *data exchange
 *trees
 *data typing
 *the same information can have different text forms and different binary
forms
 * standards

If XML can be used to encode ASN.1 documents, and if ASN.1 can be used for
schemas for XML, then the issue is not that one is better, but whether XML
_should_ well be used as a standard encoding for ASN.1 or that ASN.1 should
be the "schema" system for XML.

I don't think questions like that are solved, in the short term, on
technical reasons of capability.  People will choose their schema language
based on what they consider their most productive next step to be.  If there
are no tools, no-one will pick ASN.1.

In the long term, one would hope that the technical superiority of a
technology attracts the masses to it. If ASN.1 (or its ilk) are better,
ASN.1 fans need to figure out how to get from here to there: piggybacking on
XML is surely the only game in town. To say "Pick ASN.1 _instead_ of XML" is
to fail to learn King Canute's lesson.

Actually, it is possible to view XML's success in negative terms (I don't
really share this): that in accepting XML people are really just rejecting
CSV, flat data, undelimited data, non-web data, non-Unicode encodings, etc.
So XML was merely the first to hand for this, and could be replaced by
something superior when the clouds part and the gleaming superiority of the
other technology beams down on us all plain to see.

I would see XML's success as being rather different: I think the WWW has
shown that powerful informal systems can be built. XML's informality (whack
something out with a text editor) fits in with this.  To an extent, this
fomal/informal dichotomy (of course, there are intermediate positions) puts
XML Schemas and ASN.1 on the same side: the formal side.  I would give RSS,
SMIL, HTML as examples of informal document types: for these kinds of
things, anything more than something of the complexity of a DTD (or even a
name-spelling check) may be overkill.

Cheers
Rick Jelliffe