Title: RE: [emergency] Naming Conventions
For
Human Physical Descriptions, as pointed out in HumanML, the
public
safety industry by and large relies on standard descriptors
that
are not at the level of detail as in the medical industry. Overlaps
there
will have to be handled because the records systems for
public
safety (ie, police RMS) don't need the level of detail of
medical. I'd be surprised if that changes. That was one
reason
for the use of abstract types in the HumanML schema.
The
best we could do was pop up a metalevel and provide
supertypes, so to speak. I note that JusticeXML appears to
have
taken a similar path, and I think that is what Elliott
may be
pointing to. Naming conventions, per se, won't
fix
the problem of having two concrete types that really
are
different types.
Sorry
to belabor this. All I meant to convey was that
vocabulary based standards are reasonable, but the
more
one imports and includes the greater the likelihood
of
collisions if the domains are not aligned. It isn't always
practical to take schemas from a different industry/domain
and
attempt to factor it into another. That is why the
planning documents for a schema project of any kind
have
to be quite specific about the domain, and not,
as is
traditional in some projects, written purposefully
vague. One should not late bind domains.
Good
Friday to all.
len
-----Original Message-----
From: Rex Brooks
[mailto:]
Thanks for the input Len. Your experience is valuable, as usual. My own
concerns in this arena have been fairly limited to the overlaps with HumanML
(especially the Medical connection) and GEOVRML, and providing new kinds of
tools, i.e. 3D tools, that will use the standards this TC develops, among
others.
I had not given the database issue much thought, but what you say makes
sense.
