[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Gnumed-devel] (no subject)
From: |
catmat |
Subject: |
[Gnumed-devel] (no subject) |
Date: |
Fri, 19 Nov 2004 01:08:56 +1100 |
User-agent: |
Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.3) Gecko/20040913 |
f the clin_health_issue takes its description (name) from the most
recent episode's AOE, then the name given to the issue may have become
too narrow - maybe the student, now embarassed, is hiding a drinking
problem. Maybe clin_health_issue should just be "hypoglycemia". Maybe
the clin_health_issue, when it spans a great many episodes, needs to
be able to be given a name that communicates somethign about the
episodes as a whole, not just the AOE of the very last episode.
Another opinion :-
There's the presentation tree and the storage tree. The presentation
tree has softlinks to virtual directories (clin_health_issue) of
softlinks to (clin_root_item) , to paraphrase
a unix file directory, but the storage tree is a single rooted tree ,
with clin_root_item having clin_health_issue reference lists , but
linked to the tree via encounters.
This makes the storage tree easy to
xml-ize (or serialize if xml looks too propriety) the tree ,
and having a xsd schema description for it, means a xslt could be (?
easily) written to transform
it to something else, e.g. fragments of a hl7 message (maybe not so
easy), or html ; but no one is doing any work in this area, how
come? Is xml too propriety i.e ( another evil US plot to take over the
world by patents ) ? Python apparently has xml/xslt libraries that wrap
libxml ( a set of c libraries that are gnu licensed, I think).
Another issue, development style: waterfall, spiral, star or
rapid-prototyping /iterative ?
My feeling is that gnumed is de facto waterfall , not really spiral (
it doesn't set landmarks to reach and then move from).
It certainly more cathedral than bazaar.