RE: APIs, messaging

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

As do I.  I just find myself fighting battles 
to keep domains clean these days.  It is hard to 
do with all of these metalanguages floating around 
competing for mindshare and warping concepts 
along the way.  

On the other hand, I can build quite a bit 
of stuff and never touch XLinks (useful though 
it is).  I can build a lot less without namespaces 
because I can't aggregate in a file without it.
We only need namespace because we cracked the 
wall of the document root as named in the doctype 
and made a hierarchical namespace do a join so to speak.

This is an excellent topic for a boundary layer 
discussion.   Namespace entered the class 
definitions as a sort of "demanded hidden 
attribute value". So it's there in the grove/infoset. 
An XLink would not be in the core grove/infoset.

Let's not revisit the namespaceAsNamedSemantic 
thread today.  Nosebleed city and my anemia is 
at dangerous levels already. :-)

Separate question:  do you think the modeling language 
can warp the XML?  Do you find yourself stuffing wrapper tags 
in because: 

a) need the triple and the hierarchy gets unbalanced 
so you find yourself putting in a extra tag as switch

b) someone requires a context-free parse

In either case, something like <children> or 
<propertySets> enters an otherwise content-based 
markup language?

Len 
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: Uche Ogbuji [mailto:uche.ogbuji@fourthought.com]

Sure.  But I consider XLink a core application of XML as much as one
may consider object definition languare a core facility of ODMG.  I think
this is pretty much a matter of user preference.  After all, even
namespaces are just a layer on top.