RE: [cgmo-webcgm] QUESTION: does XCF have any ordering rules for private metadata?

["Cruikshank, David W" <david.w.cruikshank@boeing.com>]

This question did come up:

http://www.oasis-open.org/apps/org/workgroup/cgmo-webcgm/email/archives/200502/msg00036.html

and there were follow-ups from Franck and Dieter.

 

This works if the definition of the extension starts with an "|", ie <!ENTITY % grobjectEXT "|  myelement" >

Do we have any place we give guidence on how to create the extensions?  I don't remember.

 

thx...Dave

---

From: Lofton Henderson [mailto:lofton@rockynet.com]
Sent: Friday, September 23, 2005 2:07 PM
To: cgmo-webcgm@lists.oasis-open.org
Subject: [cgmo-webcgm] QUESTION: does XCF have any ordering rules for private metadata?

QUESTION:  does XCF have any ordering rules for 
application-specific metadata?

DISCUSSION:
I vaguely remember that we 
said at some time in the past, application-specific metadata elements must 
follow standardized elements.  That isn't reflected in the XCF 
anywhere.  At least I can't find it, if its there.

There are only 
two scenarios in XCF where any standardized children are allowed:

1.) 
webcgm can have any of the standardized children, grobject, layer, para, 
subpara, bindById, bindByName in any order;

2.) the only standardized 
child that any of these (except layer) can have is linkuri.

About #1 and 
metadata elements, the DTD says:
<!ELEMENT webcgm ( 
    layer | grobject | para | subpara | 
    bindById | bindByName %webcgmEXT;)* )

About 
#2 and metadata elements, the DTD says for example:
<!ELEMENT grobject 
( linkuri %grobjectEXT; )* >

So the spec says that 
application-specific metadata elements can mix with standardized elements, in 
any order.

I don't mean to raise a new issue about this, but I only want 
to clarify whether the spec accurately reflects past decisions or not.  
Does anyone recall, and preferably have a pointer to, any past element ordering 
decisions?

RECOMMENDATION:  none (just seeking clarification).