Re: [xml-dev] Holographic XML

[Stephen Green <stephengreenubl@gmail.com>]

So maybe something like this would be considered '3D' rather than '2D'
(but it might need RelaxNG):

Given a rather 2D instance or fragment

<a>
	<b>
		<c/>
		<d/>
		<e/>
	</b>
</a>


Adding some multiple occurance moves in the direction of looking more 3D

<a>
	<b>
		<c/>
		<d/>
		<e/>
	</b>	
	<b>
		<c/>
		<d/>
		<e/>
	</b>
</a>

but the occurances are only linked by having the same parent and the
exact same structure
Nevertheless there is an axis of position b[x]. Even 3D still needs
some freedom of variability
in one or more of those dimensions which accentuates that dimension; a
cylinder isn't a huge
step into the third dimension from a circle - what we need is a sphere
to make it look truly 3D.

So add some variability by allowing b to be replaced by any
'derivative' of b - 'b1' - and c with a
derivative of c (c1, etc). Also let it be allowed that b1 can be
replaced with a derivative of b1
called b2. The derivative rules might be constrained using a schema language.


<a>
	<b>
		<c/>
		<d/>
		<e/>
	</b>
	<b1>
		<c/>
		<d/>
		<e/>
		<f/>
	</b1>
	<b2>
		<c1/>
		<d/>
		<e/>
		<f/>
	</b2>
</a>

Now there is a relationship between b and b1 and between b1 and b2 in
that the structural
constraints are related (by derivation).

I guess you could say this is still 2D but you have several fragments,
each related to the
others by derivation between their structures (the constraints on
their structures) and by
their common parent and by their positions as children within the a
sequence of children.
Thus there are two orderings going on at the same time - position
(a/*[1], a/*[2], etc)
and order of derivation (b1 derives from b so it follows b in
structure, b2 derives from b1
so it follows b1 in structure). Having two axes like this is a kind of
conceptual 3D which
might be what people are after when they say some XML is too 2D.


Best regards

Steve
---
Stephen D Green



On 9 September 2010 12:21, Stephen Green <stephengreenubl@gmail.com> wrote:
>
> When we were working on the XML design of Universal Business Language at
> least one contributor thought our design was too '2-Dimensional' and we thought
> we might be able to improve things by using substitution groups (we were self-
> constrained within what could be achieved with W3C XML Schema version 1).
>
> Some thought a kind of polymorphism could be achieved this way which would
> help with customizing at schema level
> http://docs.oasis-open.org/ubl/cd-UBL-1.0/doc/cm/wd-ubl-cmsc-cmguidelines-1.0.html
>
> We abandoned this (in UBLv2) due to weaknesses found in how XML Schema 1.0's
> substitution groups worked (and were implemented) but there remains a thought
> that if we had RelaxNG we might have achieved this kind of 'holy grail' polymorphism.
> I suggest it might be worth looking at how RelaxNG might provide an alternative to
> substitution groups and allow the XML it constrains to be more '3D'.
> Best regards
>
> Steve
> ---
> Stephen D Green
>
>
>
> On 7 September 2010 16:46, Costello, Roger L. <costello@mitre.org> wrote:
>>
>> Hi Folks,
>>
>> Just brainstorming ...
>>
>> Consider:
>>
>> 1. Holography is where the information about the 3 dimensions is stored in 2 dimensional space.
>>
>> 2. I've been told by industry experts that if you're not in the 3-D memory business in four years, you're not going to be in the memory business.
>>
>> http://www.eurekalert.org/pub_releases/2010-08/ru-soc083110.php
>>
>> Operating in 3 dimensions seems to be something that will be increasingly important.
>>
>> XML is kind of a 2 dimensional representation of data. How can XML expand to 3 dimensions?
>>
>> ...
>>
>> Here are two responses from colleagues:
>>
>> ------------------------------------------
>> 3-D storage is really only about increasing the size of an array.  There's no additional complexity, only increased space organized differently.
>>
>> byte [1024][1024] twoD;
>> byte [1024][1024][1024] threeD;
>>
>> Notice that the addressing and storage method are basically the same, but "threeD" is undeniably way bigger.
>>
>> This analogy doesn't really carry over to XML.  You can't "increase the space" or density of an XML document, because it's already arbitrary according to the user's whim.  The only way I can possibly think to carry over the analogy to XML is to increase the "degrees of freedom" in an XML document by abandoning the hierarchy/rooted-tree constraint, and making it possible to represent arbitrary graphs.  (After all, trees are only a special case of graphs)  We can already do this today with RDF and other XML-serialized graph representations.  Notice that unlike 3-D storage, this introduces different semantics, not just another array dimension.  Good for some things, not for others.
>>
>> David
>>
>> ------------------------------------------
>> I agree with David.  Along the notion of additional "degrees of freedom," check out "Colorful XML: One Hierarchy Isn't Enough" by Jagadish et al.:
>>
>> http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.70.9790&rep=rep1&type=pdf
>>
>> Peter
>>
>> ...
>>
>> Ideas?
>>
>> /Roger
>>
>>
>> _______________________________________________________________________
>>
>> XML-DEV is a publicly archived, unmoderated list hosted by OASIS
>> to support XML implementation and development. To minimize
>> spam in the archives, you must subscribe before posting.
>>
>> [Un]Subscribe/change address: http://www.oasis-open.org/mlmanage/
>> Or unsubscribe: xml-dev-unsubscribe@lists.xml.org
>> subscribe: xml-dev-subscribe@lists.xml.org
>> List archive: http://lists.xml.org/archives/xml-dev/
>> List Guidelines: http://www.oasis-open.org/maillists/guidelines.php
>>
>