• From: Len Bullard <len.bullard@u...>
• To: "W. E. Perry" <wperry@f...>, xml-dev@l...
• Date: Thu, 13 Mar 2008 13:34:07 -0600

I think the key term there is 'layers'.  A mapping system such as a dynamic
weather map is also a good example.   What your concepts don't seem to
contain is a notion of a base map.  I think that is why people are
uncomfortable with it.

The reference to the character from The Prisoner is not accidental.  The
concept is that the markup only goes so far in establishing contexts, then
the content or corpus has to be examined, then the surrounding contexts such
as the two series (The Prisoner and Rumpole of the Bailey), and then some
guess work given a) the reader/processor may not have access to all of the
information (who remembers those series?), some of the information may be
wrong (spelling of proper names vary over time), and so on.  In the end,
what one has is not "the truth" but some facts with confidence metrics and
some more to test.  No end in sight.  Probably none needed.

All that said, a local agreement (base map) and some measures of success
work for most processes.  Then the humans do the rest.

len


From: W. E. Perry [mailto:wperry@f...] 
 
Would it be helpful to visualize it as Talmud, where the layers of
commentary become, as they accrete, the target of further commentary or (a
rather different point of view) new lenses through which newer commentary
sees still older content? Or as the common law, where each precedent
becomes, as above, both the target of new citation and a lens which skews
later interpretations of older cases? I have begun to think that this
ongoing creation of 'content' out of continually created meta-content is a
fundamental intellectual process, and that there is little point asking for
a defined boundary between the two which would apply to anything more that
a particular instance (as the technical distinction which you cite in the
spec is limited to an XML instance document, not to anything so ambitious
as a 'class' of such documents, however schematically congruent its
members). However, this too-facile blurring of distinction between an
instance XML document and the 'class' of documents schematically congruent
to it may be the most common failure of strict understanding in the XML
community. And why not? The very notion of naming a GI seems inherently to
imply to our minds that there are other instances just like the one being
identified which share sufficient other properties beside the name given by
the GI to all of them, and that the naming of them with that GI is in fact
the discovery that those instances taken together are a 'class'.

I am working on a project where the securities transactions chosen for a
portfolio must be shown to correspond to the defined specifics of a trading
strategy. Each such strategy has a name, which might usefully be considered
the GI in which the strategy and, beyond the strategy, the trades
undertaken in realization of that strategy are 'rooted'--effectively the
base entity of that instance of content and meta-content. But the
strategies change in all their subtle details, often daily, in direct
response to the performance of the trades rooted upon them. To model this,
it is not sufficient to maintain each day's version of the strategy as an
instance root for the trades undertaken in that strategy on that day,
because the ongoing positions generated by each of those trades will have
unique and differing histories. Part of those histories is likely to be
that each position is rebalanced, augmented or closed out at different
times in accordance with then-current versions of the underlying strategy,
or in accordance with a replacement strategy.

IMHO the only way to 'evaluate' the correspondence of a portfolio of the
moment to a strategy is to elaborate, by process, a unique instance of
semantics which is utterly specific to the particular layering of putative
content and meta-content, considering the instance at hand as a single
whole of content, if only for the sole purpose of that particular operation
of that process.

This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the sender. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail.