(dsssl) Hypothetical question on namespaces

[Trent Shipley <tcshipley@xxxxxxxxxxxxx>]

I briefly considered using TEIL for the body of my paper and DocBook for the 
bibliography.

However, I discarded that for two reasons:
1) DocBook wasn't so much closer to the ad-hoc SGML form the bibliography 
currently has.

2) Dealing with the name-space problem could be a nightmare.

===========

I'm a couple hundred pages into the DSSSL document and have looked briefly at 
Goldfarb's annotation of the SGML standard.   Neither seems to have any real 
concept of name-space although SGML does have a weird provision for 
concurrency.  

As near as I can tell the difference between name-spaces and concurency is 
that namespaces are mutually independent while concurent lexical sets can 
actually share lexical elements [namespaces never *really* share lexical 
elements, despite the existence of homonyms.]  This implies that a document 
with concurrent SGML sets will be parsed into multiple groves that can 
intersect at certain nodes.  On the other hand, if the concurrency is in the 
form of multiple namespaces the document grove must relatively well formed.  
The lexical set of the document consists of the union of non-intersecting 
lexical sets; therefore, none of the resulting sub-groves will intersect.

================

Now namespaces are an integral part of XML practice.

nsgmls claims to deal with XML; therefore, it must cope with namespaces.
OpenJade says it can't do concurrency, but it too must be able to cope with 
XML namespaces.

     How do they do this?

================

Short of editing TEIL and the relevant modules in DocBook (or switching to 
the XML versions of both, that would be cheating since it should just work) 
would there have been an elegant way to mix-and-match two or more Document 
Types

 DSSSList info and archive:  http://www.mulberrytech.com/dsssl/dssslist