Re: Distinguishing between types of address (locn)

Hello Makx

I discovered (over some long period of time) that conversations about
address can become ambiguous. There are many many tentacles to this octopus
called 'address'   :-)

It was a huge commitment by ISO Members ( national standards bodies and
other entities e.g. World Bank, UPU, etc ) to wrestle a consensus on how to
consistently engineer an agreed concept model. Before this there has been
considerable confusion, or at least regional/jurisdiction/business
differences, on how to deal with address.

In the interests of interoperability and productivity gains there is a
great need to nail-down exactly what it is as a piece of information that,
among other things, consistently communicates location, in various ways for
various user groups.

Many find it helpful to normalise the semantics within a discussion about
address e.g. I'm not sure yet what '*two different places*' means here in
this email dialogue. Does it mean two different 'addresses'? or Two
different localities / suburbs? or Two different shopping malls? Two
different delivery points (letterboxes, doors, etc)? or Two different
countries? or Two different buildings?  or Two different census
meshblocks?  Two different "addressableObjects" ??

An 'address' *may* have an association with an 'addressable object'. At a
different *point in time* the same address *may* have an association with a
different addressable object i.e. there are *state changes* with addresses.
Life-cycle information, provenance, etc, are also important
metadata/characteristics to know, record, and transact about an address.
Absolutely critical for some use-cases.

The ISO model also identifies the conceptual link between components that
make-up an address and the representation of these components as
geographical objects e.g. roads, suburbs (places?), etc ( typically
delineated within geographical information systems (GIS) ).

Richard

PS  re '*registered address*' and '*operating address*' suggests to me that
these addresses could be conceptualised/characterised as valid addresses
belonging to two different *classes of address. *


On 25 September 2015 at 19:53, Makx Dekkers <mail@makxdekkers.com> wrote:

> Richard,
>
>
>
> Yes, I understand this, even from experience. I actually happen to live in
> a house that has entrances on both sides of the block in two parallel
> streets.
>
>
>
> I was just wondering whether Pano’s question had to do with **two
> different places** associated with a legal entity, e.g. registered versus
> operating address. If I misunderstood Pano’s question, I apologise for
> creating confusion.
>
>
>
> Makx.
>
>
>
> *From:* richard.murcott@gmail.com [mailto:richard.murcott@gmail.com]
> *Sent:* 25 September 2015 03:44
> *To:* Makx Dekkers <mail@makxdekkers.com>
> *Cc:* Pano Maria <pano.maria@taxonic.com>; public-locadd@w3.org
> *Subject:* Re: Distinguishing between types of address (locn)
>
>
>
> Re: I don't understand how one site (=physical location) can have two
> different addresses, one the registration address and one the postal
> address.
>
>
>
> An addressable object may have more than one valid address, even for the
> same class of address (e.g. physical addresses). It's a common scenario. A
> simple case is where a property is situated on the corner of two addressed
> thoroughfares. It's important to identify and relate such addresses. (alias
> addresses)
>
>
>
> The semantics and models in the new ISO standard focus on sorting these
> kinds of things, and scopes numerous other complexities and nuances about
> addressing. It's easy to underestimate the complexities that arise with
> addresses. Helpfully, we now have a concept model to guide us.
>
>
>
> Richard
>
>
>
>
>
> On 24 September 2015 at 22:25, Makx Dekkers <mail@makxdekkers.com> wrote:
>
> All,
>
> >
> > > I think this is what you should do. Unless the two sites correspond to
> > > the same site.
> >
> > I don't follow. I do mean that they are one and the same site.
> > Say I have an organization O with site S1. And S1 has a registration
> address
> > Ar1 and a postal address Ap1. Would I then have to model an additional
> > instance S1'
> > representing the same site just to express the different addresses?
> >
>
> I don't understand how one site (=physical location) can have two
> different addresses, one the registration address and one the postal
> address.
> Or is the issue that an *organisation* can have a postal address that is
> different from the registration address? If that is the issue, I'd argue
> that two different addresses are associated with different physical
> locations. E.g. the physical location of a post office box is at the post
> office, not at the location where the organisation has its office.
>
> Makx.
>
>
>