Re: Canonical Path and Path Filter proposed specificatrion

[David RR Webber - XMLGlobal <Gnosis_@compuserve.com>]

Message text written by Farrukh Najmi
>recursive
structure are bad for many XML processors and as a concession we changed to
this non-recursive (non-nested way). Len, Nikola and I felt in teh
classification sub-team that we should also allow the recursive option.
This
may happen out of the Classification sub-team.<

>>>>>>>>>>>>

Farrukh,

I hope so!

I'm lost as to why recursive is anymore intensive than iteration?!?

Unless its the uncertainty on querying - but XPath solves this by
indexing the instances using thing[1], thing[2] etc to capture level.
Since our classification is a known structure - the indexing can be
implicit in peoples implementation design.

Also - seems like right now from the XML stance we've ended up
with something of an unnatural act - using attributes when we should
be using elements and as Matt points out - letting the natural 
ability of XML to represent structure as well as data points shine
thru.

Do we need here some drafts of alternate classification structure
designs?

Thanks, DW.